diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,68847 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 9831,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.774307163187583,
+      "learning_rate": 6.779661016949153e-08,
+      "loss": 2.4261,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.21884882896487,
+      "learning_rate": 1.3559322033898305e-07,
+      "loss": 2.4695,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.482975314694887,
+      "learning_rate": 2.0338983050847458e-07,
+      "loss": 1.8327,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.85451201104805,
+      "learning_rate": 2.711864406779661e-07,
+      "loss": 2.0734,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.270952415339217,
+      "learning_rate": 3.3898305084745766e-07,
+      "loss": 2.236,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.87877942813057,
+      "learning_rate": 4.0677966101694916e-07,
+      "loss": 2.2076,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.776727913056515,
+      "learning_rate": 4.745762711864407e-07,
+      "loss": 2.1196,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.25224020480987,
+      "learning_rate": 5.423728813559322e-07,
+      "loss": 2.345,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 27.33668116834911,
+      "learning_rate": 6.101694915254238e-07,
+      "loss": 2.3076,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.181907330178706,
+      "learning_rate": 6.779661016949153e-07,
+      "loss": 2.141,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.615130567617697,
+      "learning_rate": 7.457627118644069e-07,
+      "loss": 2.2639,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.579337894350388,
+      "learning_rate": 8.135593220338983e-07,
+      "loss": 2.2511,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.051745481180966,
+      "learning_rate": 8.813559322033899e-07,
+      "loss": 2.0932,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.07477432582864,
+      "learning_rate": 9.491525423728814e-07,
+      "loss": 2.1369,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.56621520785092,
+      "learning_rate": 1.016949152542373e-06,
+      "loss": 2.1692,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.349853220880675,
+      "learning_rate": 1.0847457627118644e-06,
+      "loss": 1.9588,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.978393855508065,
+      "learning_rate": 1.152542372881356e-06,
+      "loss": 1.7003,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.129076262634308,
+      "learning_rate": 1.2203389830508477e-06,
+      "loss": 1.9801,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.20744557420552,
+      "learning_rate": 1.288135593220339e-06,
+      "loss": 1.7939,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.203685104742611,
+      "learning_rate": 1.3559322033898307e-06,
+      "loss": 1.5948,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.654869222087148,
+      "learning_rate": 1.4237288135593222e-06,
+      "loss": 1.6056,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.814571724474261,
+      "learning_rate": 1.4915254237288139e-06,
+      "loss": 1.7941,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.530069326697427,
+      "learning_rate": 1.5593220338983054e-06,
+      "loss": 1.2375,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.32660567477036,
+      "learning_rate": 1.6271186440677967e-06,
+      "loss": 1.2946,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.990179490365163,
+      "learning_rate": 1.6949152542372882e-06,
+      "loss": 1.1305,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.380671795900115,
+      "learning_rate": 1.7627118644067799e-06,
+      "loss": 0.9932,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.844307063285582,
+      "learning_rate": 1.8305084745762714e-06,
+      "loss": 1.0589,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.994303435945046,
+      "learning_rate": 1.8983050847457629e-06,
+      "loss": 0.9073,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.551427417951352,
+      "learning_rate": 1.9661016949152544e-06,
+      "loss": 1.0221,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.42154176934472,
+      "learning_rate": 2.033898305084746e-06,
+      "loss": 0.9045,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6207875594724706,
+      "learning_rate": 2.1016949152542374e-06,
+      "loss": 0.8139,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1824465930868917,
+      "learning_rate": 2.169491525423729e-06,
+      "loss": 0.669,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.706012292777276,
+      "learning_rate": 2.2372881355932204e-06,
+      "loss": 0.8022,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8251403517668487,
+      "learning_rate": 2.305084745762712e-06,
+      "loss": 0.8254,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.984073023660238,
+      "learning_rate": 2.372881355932204e-06,
+      "loss": 0.9779,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4188064951429475,
+      "learning_rate": 2.4406779661016953e-06,
+      "loss": 0.7361,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3516733355792843,
+      "learning_rate": 2.5084745762711864e-06,
+      "loss": 0.8315,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2496309758042417,
+      "learning_rate": 2.576271186440678e-06,
+      "loss": 0.8022,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.420808155220749,
+      "learning_rate": 2.64406779661017e-06,
+      "loss": 0.8303,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.235898259469345,
+      "learning_rate": 2.7118644067796613e-06,
+      "loss": 0.8828,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.259722995408095,
+      "learning_rate": 2.779661016949153e-06,
+      "loss": 0.7345,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5220575903145717,
+      "learning_rate": 2.8474576271186443e-06,
+      "loss": 0.8884,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.01432745835319,
+      "learning_rate": 2.915254237288136e-06,
+      "loss": 0.716,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3912698546304783,
+      "learning_rate": 2.9830508474576277e-06,
+      "loss": 0.8186,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2176167549704413,
+      "learning_rate": 3.0508474576271192e-06,
+      "loss": 0.8315,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3542559647297248,
+      "learning_rate": 3.1186440677966107e-06,
+      "loss": 0.8938,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.408850666287898,
+      "learning_rate": 3.186440677966102e-06,
+      "loss": 0.7443,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9403993418379064,
+      "learning_rate": 3.2542372881355933e-06,
+      "loss": 0.7316,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2266120044077855,
+      "learning_rate": 3.322033898305085e-06,
+      "loss": 0.823,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2149796287947052,
+      "learning_rate": 3.3898305084745763e-06,
+      "loss": 0.8493,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.964248261933294,
+      "learning_rate": 3.457627118644068e-06,
+      "loss": 0.7657,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1369395484319402,
+      "learning_rate": 3.5254237288135597e-06,
+      "loss": 0.7717,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9324624073386334,
+      "learning_rate": 3.5932203389830512e-06,
+      "loss": 0.7885,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7645971001845546,
+      "learning_rate": 3.6610169491525427e-06,
+      "loss": 0.6827,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9794442176736449,
+      "learning_rate": 3.7288135593220342e-06,
+      "loss": 0.7341,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.205253413175247,
+      "learning_rate": 3.7966101694915257e-06,
+      "loss": 0.8477,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9368774918084077,
+      "learning_rate": 3.864406779661018e-06,
+      "loss": 0.7182,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8331552352287894,
+      "learning_rate": 3.932203389830509e-06,
+      "loss": 0.6232,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1664795950928943,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7919,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9510344143430243,
+      "learning_rate": 4.067796610169492e-06,
+      "loss": 0.7199,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1394368935709833,
+      "learning_rate": 4.135593220338983e-06,
+      "loss": 0.846,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9579721708838056,
+      "learning_rate": 4.203389830508475e-06,
+      "loss": 0.7141,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.257998109160227,
+      "learning_rate": 4.271186440677967e-06,
+      "loss": 0.7374,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9561971563274394,
+      "learning_rate": 4.338983050847458e-06,
+      "loss": 0.6967,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.961780054829146,
+      "learning_rate": 4.40677966101695e-06,
+      "loss": 0.8134,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9708755542520096,
+      "learning_rate": 4.474576271186441e-06,
+      "loss": 0.6124,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1606992603731294,
+      "learning_rate": 4.542372881355933e-06,
+      "loss": 0.729,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0759273723089726,
+      "learning_rate": 4.610169491525424e-06,
+      "loss": 0.7483,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.19293232912226,
+      "learning_rate": 4.677966101694916e-06,
+      "loss": 0.7522,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8359859772491813,
+      "learning_rate": 4.745762711864408e-06,
+      "loss": 0.7899,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3845629041250707,
+      "learning_rate": 4.813559322033899e-06,
+      "loss": 0.7246,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1085590957497846,
+      "learning_rate": 4.881355932203391e-06,
+      "loss": 0.7739,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8918037708327846,
+      "learning_rate": 4.949152542372882e-06,
+      "loss": 0.7698,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0067874509675687,
+      "learning_rate": 5.016949152542373e-06,
+      "loss": 0.7346,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.119502972604516,
+      "learning_rate": 5.084745762711865e-06,
+      "loss": 0.8479,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.861663786810975,
+      "learning_rate": 5.152542372881356e-06,
+      "loss": 0.6784,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9695486836646774,
+      "learning_rate": 5.220338983050848e-06,
+      "loss": 0.7107,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.171762088984204,
+      "learning_rate": 5.28813559322034e-06,
+      "loss": 0.7572,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1380903015319648,
+      "learning_rate": 5.355932203389831e-06,
+      "loss": 0.7518,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.020115592796084,
+      "learning_rate": 5.423728813559323e-06,
+      "loss": 0.7172,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3246993299029914,
+      "learning_rate": 5.491525423728814e-06,
+      "loss": 0.7228,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0860284304875973,
+      "learning_rate": 5.559322033898306e-06,
+      "loss": 0.7337,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.105468963098628,
+      "learning_rate": 5.6271186440677975e-06,
+      "loss": 0.8502,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9810307968538619,
+      "learning_rate": 5.694915254237289e-06,
+      "loss": 0.7639,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.42547864581517,
+      "learning_rate": 5.7627118644067805e-06,
+      "loss": 0.8198,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9843324749634157,
+      "learning_rate": 5.830508474576272e-06,
+      "loss": 0.6272,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8295596514924568,
+      "learning_rate": 5.8983050847457635e-06,
+      "loss": 0.6614,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2018819268618697,
+      "learning_rate": 5.9661016949152555e-06,
+      "loss": 0.8396,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.014841623849869,
+      "learning_rate": 6.0338983050847465e-06,
+      "loss": 0.7825,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1391290710966953,
+      "learning_rate": 6.1016949152542385e-06,
+      "loss": 0.6737,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.844927353698065,
+      "learning_rate": 6.1694915254237295e-06,
+      "loss": 0.7386,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4827254198905573,
+      "learning_rate": 6.2372881355932215e-06,
+      "loss": 0.9183,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1865083435662775,
+      "learning_rate": 6.3050847457627125e-06,
+      "loss": 0.6657,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.22914258959008,
+      "learning_rate": 6.372881355932204e-06,
+      "loss": 0.7878,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0465478945502737,
+      "learning_rate": 6.440677966101695e-06,
+      "loss": 0.7384,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.150182007226782,
+      "learning_rate": 6.508474576271187e-06,
+      "loss": 0.7675,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.20138713129849,
+      "learning_rate": 6.576271186440678e-06,
+      "loss": 0.7957,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1717654881518405,
+      "learning_rate": 6.64406779661017e-06,
+      "loss": 0.7656,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1766599305935683,
+      "learning_rate": 6.7118644067796615e-06,
+      "loss": 0.8074,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.250178867344064,
+      "learning_rate": 6.779661016949153e-06,
+      "loss": 0.7486,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1036428186662457,
+      "learning_rate": 6.8474576271186445e-06,
+      "loss": 0.7276,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.231120104885962,
+      "learning_rate": 6.915254237288136e-06,
+      "loss": 0.7694,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8320331250168351,
+      "learning_rate": 6.9830508474576275e-06,
+      "loss": 0.7384,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.07040534417232,
+      "learning_rate": 7.0508474576271195e-06,
+      "loss": 0.6566,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.052812058503903,
+      "learning_rate": 7.1186440677966106e-06,
+      "loss": 0.761,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1483979394263417,
+      "learning_rate": 7.1864406779661025e-06,
+      "loss": 0.7733,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.029441020419923,
+      "learning_rate": 7.2542372881355936e-06,
+      "loss": 0.6775,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0568628127678794,
+      "learning_rate": 7.3220338983050855e-06,
+      "loss": 0.6425,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.179569873526594,
+      "learning_rate": 7.3898305084745766e-06,
+      "loss": 0.7068,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9555856066898845,
+      "learning_rate": 7.4576271186440685e-06,
+      "loss": 0.7595,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3202948797607554,
+      "learning_rate": 7.52542372881356e-06,
+      "loss": 0.8602,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0643122915003107,
+      "learning_rate": 7.5932203389830515e-06,
+      "loss": 0.791,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2126598533146677,
+      "learning_rate": 7.661016949152543e-06,
+      "loss": 0.7064,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4298478880835463,
+      "learning_rate": 7.728813559322035e-06,
+      "loss": 0.7869,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.288531605902557,
+      "learning_rate": 7.796610169491526e-06,
+      "loss": 0.896,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2745017660805855,
+      "learning_rate": 7.864406779661017e-06,
+      "loss": 0.6802,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6570737256402928,
+      "learning_rate": 7.93220338983051e-06,
+      "loss": 0.6276,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1733978454998923,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7266,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.142950183320776,
+      "learning_rate": 8.067796610169492e-06,
+      "loss": 0.7994,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1131892760021556,
+      "learning_rate": 8.135593220338983e-06,
+      "loss": 0.7306,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9104523566550864,
+      "learning_rate": 8.203389830508475e-06,
+      "loss": 0.7925,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9144516822731146,
+      "learning_rate": 8.271186440677966e-06,
+      "loss": 0.7236,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1479589283189173,
+      "learning_rate": 8.338983050847458e-06,
+      "loss": 0.7482,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.154494552349858,
+      "learning_rate": 8.40677966101695e-06,
+      "loss": 0.689,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9714017351330468,
+      "learning_rate": 8.47457627118644e-06,
+      "loss": 0.7369,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.988551787371157,
+      "learning_rate": 8.542372881355933e-06,
+      "loss": 0.7643,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.325041578076891,
+      "learning_rate": 8.610169491525424e-06,
+      "loss": 0.7964,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.176814754004798,
+      "learning_rate": 8.677966101694915e-06,
+      "loss": 0.7643,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0824812157544645,
+      "learning_rate": 8.745762711864407e-06,
+      "loss": 0.7471,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9684242759673636,
+      "learning_rate": 8.8135593220339e-06,
+      "loss": 0.8099,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0678745627639197,
+      "learning_rate": 8.88135593220339e-06,
+      "loss": 0.8663,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.060252010222962,
+      "learning_rate": 8.949152542372881e-06,
+      "loss": 0.7509,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8677177304612413,
+      "learning_rate": 9.016949152542374e-06,
+      "loss": 0.719,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0650299093701965,
+      "learning_rate": 9.084745762711865e-06,
+      "loss": 0.6587,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.157297646435543,
+      "learning_rate": 9.152542372881356e-06,
+      "loss": 0.7644,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9208876885040105,
+      "learning_rate": 9.220338983050847e-06,
+      "loss": 0.7299,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1636200591275476,
+      "learning_rate": 9.28813559322034e-06,
+      "loss": 0.8205,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.240165955523122,
+      "learning_rate": 9.355932203389831e-06,
+      "loss": 0.7958,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3038562909608458,
+      "learning_rate": 9.423728813559322e-06,
+      "loss": 0.7342,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8045909570080936,
+      "learning_rate": 9.491525423728815e-06,
+      "loss": 0.6016,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8861330770480502,
+      "learning_rate": 9.559322033898306e-06,
+      "loss": 0.7197,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.973267688604536,
+      "learning_rate": 9.627118644067797e-06,
+      "loss": 0.7593,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8698474554660958,
+      "learning_rate": 9.69491525423729e-06,
+      "loss": 0.6507,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0010518197546987,
+      "learning_rate": 9.762711864406781e-06,
+      "loss": 0.7075,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9194243510439397,
+      "learning_rate": 9.830508474576272e-06,
+      "loss": 0.802,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.168846656888089,
+      "learning_rate": 9.898305084745763e-06,
+      "loss": 0.7935,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.125833442803335,
+      "learning_rate": 9.966101694915256e-06,
+      "loss": 0.7864,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.102996254972869,
+      "learning_rate": 1.0033898305084746e-05,
+      "loss": 0.7078,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1987112323642615,
+      "learning_rate": 1.0101694915254238e-05,
+      "loss": 0.7534,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.815570288794822,
+      "learning_rate": 1.016949152542373e-05,
+      "loss": 0.8229,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1419109745500293,
+      "learning_rate": 1.0237288135593222e-05,
+      "loss": 0.7117,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1575782595640476,
+      "learning_rate": 1.0305084745762712e-05,
+      "loss": 0.8649,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3369718017851002,
+      "learning_rate": 1.0372881355932204e-05,
+      "loss": 0.841,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.89934679393495,
+      "learning_rate": 1.0440677966101695e-05,
+      "loss": 0.6265,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.116565213986925,
+      "learning_rate": 1.0508474576271188e-05,
+      "loss": 0.724,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.982021753415292,
+      "learning_rate": 1.057627118644068e-05,
+      "loss": 0.789,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1139135488892666,
+      "learning_rate": 1.0644067796610172e-05,
+      "loss": 0.8074,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.118524566834072,
+      "learning_rate": 1.0711864406779661e-05,
+      "loss": 0.7203,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.12820506314938,
+      "learning_rate": 1.0779661016949154e-05,
+      "loss": 0.7984,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9423385020630755,
+      "learning_rate": 1.0847457627118645e-05,
+      "loss": 0.6801,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1069618355533097,
+      "learning_rate": 1.0915254237288135e-05,
+      "loss": 0.6958,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2371552107869843,
+      "learning_rate": 1.0983050847457627e-05,
+      "loss": 0.5289,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.904024031354274,
+      "learning_rate": 1.1050847457627118e-05,
+      "loss": 0.6713,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.127282166318957,
+      "learning_rate": 1.1118644067796611e-05,
+      "loss": 0.7227,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.142497901612626,
+      "learning_rate": 1.1186440677966102e-05,
+      "loss": 0.5906,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.057330453800443,
+      "learning_rate": 1.1254237288135595e-05,
+      "loss": 0.6901,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.093010939189739,
+      "learning_rate": 1.1322033898305084e-05,
+      "loss": 0.7951,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9107896944652738,
+      "learning_rate": 1.1389830508474577e-05,
+      "loss": 0.7762,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.27565600735753,
+      "learning_rate": 1.1457627118644068e-05,
+      "loss": 0.7201,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.078256587268634,
+      "learning_rate": 1.1525423728813561e-05,
+      "loss": 0.7271,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.976498895581692,
+      "learning_rate": 1.159322033898305e-05,
+      "loss": 0.7205,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0274252035867497,
+      "learning_rate": 1.1661016949152543e-05,
+      "loss": 0.7542,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.248882635591796,
+      "learning_rate": 1.1728813559322034e-05,
+      "loss": 0.7563,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.152946416343347,
+      "learning_rate": 1.1796610169491527e-05,
+      "loss": 0.8891,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9614188493029023,
+      "learning_rate": 1.1864406779661018e-05,
+      "loss": 0.6927,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9561016698575437,
+      "learning_rate": 1.1932203389830511e-05,
+      "loss": 0.7323,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2060339927740333,
+      "learning_rate": 1.2e-05,
+      "loss": 0.7635,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.361699969994298,
+      "learning_rate": 1.2067796610169493e-05,
+      "loss": 0.7074,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2826526992705576,
+      "learning_rate": 1.2135593220338984e-05,
+      "loss": 0.8152,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3383409650671783,
+      "learning_rate": 1.2203389830508477e-05,
+      "loss": 0.7099,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.314928794545433,
+      "learning_rate": 1.2271186440677966e-05,
+      "loss": 0.8114,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1182102763924564,
+      "learning_rate": 1.2338983050847459e-05,
+      "loss": 0.7584,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9427488854499315,
+      "learning_rate": 1.240677966101695e-05,
+      "loss": 0.6497,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.044035298732104,
+      "learning_rate": 1.2474576271186443e-05,
+      "loss": 0.8107,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.000921224831445,
+      "learning_rate": 1.2542372881355932e-05,
+      "loss": 0.7334,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.148842701328798,
+      "learning_rate": 1.2610169491525425e-05,
+      "loss": 0.8103,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2665184439248676,
+      "learning_rate": 1.2677966101694916e-05,
+      "loss": 0.8093,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9357363672299526,
+      "learning_rate": 1.2745762711864407e-05,
+      "loss": 0.7027,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7570604626310313,
+      "learning_rate": 1.28135593220339e-05,
+      "loss": 0.6212,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2773162054308607,
+      "learning_rate": 1.288135593220339e-05,
+      "loss": 0.7666,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7761063488705422,
+      "learning_rate": 1.2949152542372882e-05,
+      "loss": 0.6996,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8767194093067885,
+      "learning_rate": 1.3016949152542373e-05,
+      "loss": 0.7106,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0091109182168436,
+      "learning_rate": 1.3084745762711866e-05,
+      "loss": 0.6987,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8537469691014754,
+      "learning_rate": 1.3152542372881355e-05,
+      "loss": 0.6861,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.017545763530024,
+      "learning_rate": 1.3220338983050848e-05,
+      "loss": 0.7587,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1514998030012236,
+      "learning_rate": 1.328813559322034e-05,
+      "loss": 0.828,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9143327758731226,
+      "learning_rate": 1.3355932203389832e-05,
+      "loss": 0.7598,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9042594463315574,
+      "learning_rate": 1.3423728813559323e-05,
+      "loss": 0.7754,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8556994255207242,
+      "learning_rate": 1.3491525423728816e-05,
+      "loss": 0.7414,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1774086553474232,
+      "learning_rate": 1.3559322033898305e-05,
+      "loss": 0.7429,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9337554264476329,
+      "learning_rate": 1.3627118644067798e-05,
+      "loss": 0.6986,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.718424449386759,
+      "learning_rate": 1.3694915254237289e-05,
+      "loss": 0.68,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.906706540088502,
+      "learning_rate": 1.3762711864406782e-05,
+      "loss": 0.822,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.098824628782942,
+      "learning_rate": 1.3830508474576271e-05,
+      "loss": 0.7159,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.798103139341794,
+      "learning_rate": 1.3898305084745764e-05,
+      "loss": 0.817,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6708618914113391,
+      "learning_rate": 1.3966101694915255e-05,
+      "loss": 0.7206,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8960631122303189,
+      "learning_rate": 1.4033898305084748e-05,
+      "loss": 0.7623,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.941715727270628,
+      "learning_rate": 1.4101694915254239e-05,
+      "loss": 0.664,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.964233432476786,
+      "learning_rate": 1.416949152542373e-05,
+      "loss": 0.6237,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7513382787513723,
+      "learning_rate": 1.4237288135593221e-05,
+      "loss": 0.6544,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9443552456303226,
+      "learning_rate": 1.4305084745762714e-05,
+      "loss": 0.6672,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9412006314237946,
+      "learning_rate": 1.4372881355932205e-05,
+      "loss": 0.8645,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.117636061000563,
+      "learning_rate": 1.4440677966101698e-05,
+      "loss": 0.686,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0391785941599845,
+      "learning_rate": 1.4508474576271187e-05,
+      "loss": 0.8448,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.950652111059504,
+      "learning_rate": 1.4576271186440678e-05,
+      "loss": 0.6526,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.987977720735387,
+      "learning_rate": 1.4644067796610171e-05,
+      "loss": 0.7114,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.173671919832719,
+      "learning_rate": 1.4711864406779662e-05,
+      "loss": 0.706,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0170009363485395,
+      "learning_rate": 1.4779661016949153e-05,
+      "loss": 0.6986,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7783761996339809,
+      "learning_rate": 1.4847457627118644e-05,
+      "loss": 0.7819,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.120675334171391,
+      "learning_rate": 1.4915254237288137e-05,
+      "loss": 0.7375,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8559328669266935,
+      "learning_rate": 1.4983050847457628e-05,
+      "loss": 0.7511,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9057535498740088,
+      "learning_rate": 1.505084745762712e-05,
+      "loss": 0.7453,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9176427912162302,
+      "learning_rate": 1.511864406779661e-05,
+      "loss": 0.65,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.967862907768235,
+      "learning_rate": 1.5186440677966103e-05,
+      "loss": 0.7469,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8613239875337846,
+      "learning_rate": 1.5254237288135594e-05,
+      "loss": 0.8355,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6798452172801586,
+      "learning_rate": 1.5322033898305085e-05,
+      "loss": 0.7247,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9657751054133528,
+      "learning_rate": 1.5389830508474578e-05,
+      "loss": 0.6851,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8171542649675818,
+      "learning_rate": 1.545762711864407e-05,
+      "loss": 0.7742,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.010455167366286,
+      "learning_rate": 1.552542372881356e-05,
+      "loss": 0.8376,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.891522214459134,
+      "learning_rate": 1.5593220338983053e-05,
+      "loss": 0.8006,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8244670277804427,
+      "learning_rate": 1.5661016949152542e-05,
+      "loss": 0.6862,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.035296790223969,
+      "learning_rate": 1.5728813559322035e-05,
+      "loss": 0.7342,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.036716471677796,
+      "learning_rate": 1.5796610169491528e-05,
+      "loss": 0.7172,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.966447074428311,
+      "learning_rate": 1.586440677966102e-05,
+      "loss": 0.778,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8457730222414168,
+      "learning_rate": 1.593220338983051e-05,
+      "loss": 0.628,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8050930389522784,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7722,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.531322108969687,
+      "learning_rate": 1.6067796610169492e-05,
+      "loss": 0.8066,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0927576851100174,
+      "learning_rate": 1.6135593220338985e-05,
+      "loss": 0.7249,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.029523839237554,
+      "learning_rate": 1.6203389830508474e-05,
+      "loss": 0.8032,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0478471468460615,
+      "learning_rate": 1.6271186440677967e-05,
+      "loss": 0.7326,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.829411642842363,
+      "learning_rate": 1.633898305084746e-05,
+      "loss": 0.817,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8488495872202233,
+      "learning_rate": 1.640677966101695e-05,
+      "loss": 0.6522,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9832710283695125,
+      "learning_rate": 1.6474576271186442e-05,
+      "loss": 0.7512,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.100562680767454,
+      "learning_rate": 1.654237288135593e-05,
+      "loss": 0.8037,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.035367004658631,
+      "learning_rate": 1.6610169491525424e-05,
+      "loss": 0.8428,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7483550173167273,
+      "learning_rate": 1.6677966101694917e-05,
+      "loss": 0.75,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.930898045070027,
+      "learning_rate": 1.674576271186441e-05,
+      "loss": 0.6774,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8365756306413494,
+      "learning_rate": 1.68135593220339e-05,
+      "loss": 0.723,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.701138383260526,
+      "learning_rate": 1.6881355932203392e-05,
+      "loss": 0.7625,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7105626859044865,
+      "learning_rate": 1.694915254237288e-05,
+      "loss": 0.6939,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.059412608350948,
+      "learning_rate": 1.7016949152542374e-05,
+      "loss": 0.6739,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.712149442270916,
+      "learning_rate": 1.7084745762711867e-05,
+      "loss": 0.744,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.015776887334701,
+      "learning_rate": 1.715254237288136e-05,
+      "loss": 0.7223,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.685804526129291,
+      "learning_rate": 1.722033898305085e-05,
+      "loss": 0.7029,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9446296967732069,
+      "learning_rate": 1.728813559322034e-05,
+      "loss": 0.7818,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.791749565441212,
+      "learning_rate": 1.735593220338983e-05,
+      "loss": 0.6184,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7154113226867052,
+      "learning_rate": 1.7423728813559324e-05,
+      "loss": 0.7036,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.883567618364835,
+      "learning_rate": 1.7491525423728813e-05,
+      "loss": 0.7517,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8940636144580816,
+      "learning_rate": 1.7559322033898306e-05,
+      "loss": 0.7078,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9268538873409744,
+      "learning_rate": 1.76271186440678e-05,
+      "loss": 0.7563,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7748203123085324,
+      "learning_rate": 1.769491525423729e-05,
+      "loss": 0.7079,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.76578312180332,
+      "learning_rate": 1.776271186440678e-05,
+      "loss": 0.6875,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8139793782816562,
+      "learning_rate": 1.7830508474576274e-05,
+      "loss": 0.7254,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7496909186973184,
+      "learning_rate": 1.7898305084745763e-05,
+      "loss": 0.7714,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9619860178080115,
+      "learning_rate": 1.7966101694915256e-05,
+      "loss": 0.6743,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9633822472809355,
+      "learning_rate": 1.803389830508475e-05,
+      "loss": 0.7198,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8938463788084676,
+      "learning_rate": 1.810169491525424e-05,
+      "loss": 0.8072,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7910793721294165,
+      "learning_rate": 1.816949152542373e-05,
+      "loss": 0.7417,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8385244007491037,
+      "learning_rate": 1.823728813559322e-05,
+      "loss": 0.7399,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8299310983186288,
+      "learning_rate": 1.8305084745762713e-05,
+      "loss": 0.8197,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6938430742018853,
+      "learning_rate": 1.8372881355932202e-05,
+      "loss": 0.7432,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.97932361867903,
+      "learning_rate": 1.8440677966101695e-05,
+      "loss": 0.7969,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7016617079758896,
+      "learning_rate": 1.8508474576271188e-05,
+      "loss": 0.7311,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7402288852531422,
+      "learning_rate": 1.857627118644068e-05,
+      "loss": 0.8583,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8009302958397473,
+      "learning_rate": 1.864406779661017e-05,
+      "loss": 0.7303,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8701338309527233,
+      "learning_rate": 1.8711864406779663e-05,
+      "loss": 0.7549,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0312194476891876,
+      "learning_rate": 1.8779661016949152e-05,
+      "loss": 0.7222,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8677846365652164,
+      "learning_rate": 1.8847457627118645e-05,
+      "loss": 0.7295,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.949694415094439,
+      "learning_rate": 1.8915254237288138e-05,
+      "loss": 0.7984,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7111163200955561,
+      "learning_rate": 1.898305084745763e-05,
+      "loss": 0.7203,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9717239024751367,
+      "learning_rate": 1.905084745762712e-05,
+      "loss": 0.6803,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.579686370395743,
+      "learning_rate": 1.9118644067796613e-05,
+      "loss": 0.7824,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.678149645995694,
+      "learning_rate": 1.9186440677966102e-05,
+      "loss": 0.6746,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8213188558433615,
+      "learning_rate": 1.9254237288135595e-05,
+      "loss": 0.8051,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7725977550117176,
+      "learning_rate": 1.9322033898305087e-05,
+      "loss": 0.7281,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9372054768281428,
+      "learning_rate": 1.938983050847458e-05,
+      "loss": 0.6533,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7535765837740318,
+      "learning_rate": 1.945762711864407e-05,
+      "loss": 0.8288,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.859536473734018,
+      "learning_rate": 1.9525423728813562e-05,
+      "loss": 0.8189,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.843961540210964,
+      "learning_rate": 1.9593220338983052e-05,
+      "loss": 0.7668,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8461464381997479,
+      "learning_rate": 1.9661016949152545e-05,
+      "loss": 0.7714,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7238192387959088,
+      "learning_rate": 1.9728813559322034e-05,
+      "loss": 0.7324,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6427586720320793,
+      "learning_rate": 1.9796610169491527e-05,
+      "loss": 0.7335,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6800342906853636,
+      "learning_rate": 1.986440677966102e-05,
+      "loss": 0.7549,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7832389725065179,
+      "learning_rate": 1.9932203389830512e-05,
+      "loss": 0.8058,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.727775985861393,
+      "learning_rate": 2e-05,
+      "loss": 0.6759,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7977238751907165,
+      "learning_rate": 1.999999945732819e-05,
+      "loss": 0.7674,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7390933788321141,
+      "learning_rate": 1.9999997829312825e-05,
+      "loss": 0.8395,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7764290549311015,
+      "learning_rate": 1.9999995115954075e-05,
+      "loss": 0.711,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6826307129531393,
+      "learning_rate": 1.999999131725224e-05,
+      "loss": 0.6686,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.664467172792983,
+      "learning_rate": 1.9999986433207727e-05,
+      "loss": 0.6859,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.91798902000967,
+      "learning_rate": 1.9999980463821073e-05,
+      "loss": 0.7042,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6449618209624421,
+      "learning_rate": 1.9999973409092916e-05,
+      "loss": 0.6718,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7583224606564245,
+      "learning_rate": 1.999996526902403e-05,
+      "loss": 0.7526,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7418882410175618,
+      "learning_rate": 1.99999560436153e-05,
+      "loss": 0.8077,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6483278016364271,
+      "learning_rate": 1.9999945732867717e-05,
+      "loss": 0.6834,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8758827011232249,
+      "learning_rate": 1.999993433678241e-05,
+      "loss": 0.6097,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9950744094762918,
+      "learning_rate": 1.999992185536061e-05,
+      "loss": 0.8323,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9600308793601466,
+      "learning_rate": 1.9999908288603678e-05,
+      "loss": 0.7898,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6362825731715926,
+      "learning_rate": 1.9999893636513078e-05,
+      "loss": 0.715,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9191632580569564,
+      "learning_rate": 1.9999877899090408e-05,
+      "loss": 0.8983,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2077252767571283,
+      "learning_rate": 1.999986107633737e-05,
+      "loss": 0.8502,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7211750642056591,
+      "learning_rate": 1.9999843168255792e-05,
+      "loss": 0.6766,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8405066522264266,
+      "learning_rate": 1.999982417484762e-05,
+      "loss": 0.7717,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7584962556190842,
+      "learning_rate": 1.9999804096114914e-05,
+      "loss": 0.7634,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9067370561909576,
+      "learning_rate": 1.999978293205985e-05,
+      "loss": 0.7854,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.650263611438605,
+      "learning_rate": 1.9999760682684732e-05,
+      "loss": 0.8554,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9239042342081574,
+      "learning_rate": 1.999973734799197e-05,
+      "loss": 0.7182,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.914889513011852,
+      "learning_rate": 1.9999712927984097e-05,
+      "loss": 0.8431,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7596818561456582,
+      "learning_rate": 1.999968742266376e-05,
+      "loss": 0.7526,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9152360275383022,
+      "learning_rate": 1.9999660832033733e-05,
+      "loss": 0.8691,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7674570266140242,
+      "learning_rate": 1.99996331560969e-05,
+      "loss": 0.7009,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8346671747992218,
+      "learning_rate": 1.9999604394856265e-05,
+      "loss": 0.7458,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7024441673305173,
+      "learning_rate": 1.999957454831495e-05,
+      "loss": 0.6926,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8858862630231061,
+      "learning_rate": 1.9999543616476195e-05,
+      "loss": 0.7739,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.918639765230697,
+      "learning_rate": 1.999951159934335e-05,
+      "loss": 0.7835,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.713705802835458,
+      "learning_rate": 1.99994784969199e-05,
+      "loss": 0.8373,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.125556618290798,
+      "learning_rate": 1.9999444309209432e-05,
+      "loss": 0.7237,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8747153782294466,
+      "learning_rate": 1.999940903621566e-05,
+      "loss": 0.8242,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8541269896488926,
+      "learning_rate": 1.999937267794241e-05,
+      "loss": 0.8078,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8097939569533714,
+      "learning_rate": 1.999933523439363e-05,
+      "loss": 0.6798,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6707764941842447,
+      "learning_rate": 1.999929670557338e-05,
+      "loss": 0.7481,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5408400944729324,
+      "learning_rate": 1.9999257091485842e-05,
+      "loss": 0.6834,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8105877361881173,
+      "learning_rate": 1.999921639213532e-05,
+      "loss": 0.7798,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0245928171672127,
+      "learning_rate": 1.9999174607526232e-05,
+      "loss": 0.8071,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7126779698650814,
+      "learning_rate": 1.9999131737663106e-05,
+      "loss": 0.7318,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8799372221054982,
+      "learning_rate": 1.9999087782550596e-05,
+      "loss": 0.7915,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8302157825287524,
+      "learning_rate": 1.999904274219348e-05,
+      "loss": 0.8132,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8873097603497921,
+      "learning_rate": 1.9998996616596643e-05,
+      "loss": 0.818,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7616359225670495,
+      "learning_rate": 1.9998949405765086e-05,
+      "loss": 0.6288,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9961891473127835,
+      "learning_rate": 1.9998901109703942e-05,
+      "loss": 0.7461,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9590109916459193,
+      "learning_rate": 1.9998851728418443e-05,
+      "loss": 0.7262,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7017357772857418,
+      "learning_rate": 1.9998801261913956e-05,
+      "loss": 0.697,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8180055536057091,
+      "learning_rate": 1.9998749710195957e-05,
+      "loss": 0.7893,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0549880336112016,
+      "learning_rate": 1.9998697073270038e-05,
+      "loss": 0.7717,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6411992346494544,
+      "learning_rate": 1.9998643351141912e-05,
+      "loss": 0.7945,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8553473962687779,
+      "learning_rate": 1.9998588543817415e-05,
+      "loss": 0.8276,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8223722140799334,
+      "learning_rate": 1.999853265130249e-05,
+      "loss": 0.842,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7858229071600773,
+      "learning_rate": 1.9998475673603205e-05,
+      "loss": 0.6635,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5519682746904044,
+      "learning_rate": 1.9998417610725745e-05,
+      "loss": 0.6665,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9122831115958798,
+      "learning_rate": 1.999835846267641e-05,
+      "loss": 0.7812,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7934216535596592,
+      "learning_rate": 1.9998298229461624e-05,
+      "loss": 0.7411,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.714151647311422,
+      "learning_rate": 1.999823691108792e-05,
+      "loss": 0.7088,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8445691735054452,
+      "learning_rate": 1.9998174507561952e-05,
+      "loss": 0.8371,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.946917458223072,
+      "learning_rate": 1.9998111018890496e-05,
+      "loss": 0.8158,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9627402550609958,
+      "learning_rate": 1.999804644508044e-05,
+      "loss": 0.816,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.684490996381524,
+      "learning_rate": 1.99979807861388e-05,
+      "loss": 0.7906,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7292982143939393,
+      "learning_rate": 1.999791404207269e-05,
+      "loss": 0.8777,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6990863259567761,
+      "learning_rate": 1.9997846212889367e-05,
+      "loss": 0.7051,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6806371114663146,
+      "learning_rate": 1.999777729859618e-05,
+      "loss": 0.6675,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.891811437481674,
+      "learning_rate": 1.9997707299200622e-05,
+      "loss": 0.8882,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7243799607340802,
+      "learning_rate": 1.999763621471028e-05,
+      "loss": 0.7244,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8932374823398974,
+      "learning_rate": 1.999756404513287e-05,
+      "loss": 0.7592,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7673934736131074,
+      "learning_rate": 1.999749079047623e-05,
+      "loss": 0.7581,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.688678447260584,
+      "learning_rate": 1.9997416450748308e-05,
+      "loss": 0.7357,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8030241010019676,
+      "learning_rate": 1.999734102595717e-05,
+      "loss": 0.8179,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6616197218353204,
+      "learning_rate": 1.9997264516111006e-05,
+      "loss": 0.6571,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7328480272335298,
+      "learning_rate": 1.9997186921218123e-05,
+      "loss": 0.8508,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.591325686668673,
+      "learning_rate": 1.999710824128693e-05,
+      "loss": 0.7019,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7290334221881392,
+      "learning_rate": 1.9997028476325978e-05,
+      "loss": 0.8135,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9651039231608671,
+      "learning_rate": 1.9996947626343924e-05,
+      "loss": 0.8036,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8616771729798702,
+      "learning_rate": 1.9996865691349538e-05,
+      "loss": 0.7986,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.656935611345808,
+      "learning_rate": 1.9996782671351713e-05,
+      "loss": 0.7317,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8377731073141668,
+      "learning_rate": 1.9996698566359462e-05,
+      "loss": 0.7378,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8112813049650685,
+      "learning_rate": 1.9996613376381916e-05,
+      "loss": 0.7948,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0118770931741934,
+      "learning_rate": 1.999652710142831e-05,
+      "loss": 0.8042,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7478808769772007,
+      "learning_rate": 1.999643974150802e-05,
+      "loss": 0.7676,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7676338042836226,
+      "learning_rate": 1.9996351296630525e-05,
+      "loss": 0.7365,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7180183778756248,
+      "learning_rate": 1.999626176680542e-05,
+      "loss": 0.6893,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.50083822299388,
+      "learning_rate": 1.9996171152042425e-05,
+      "loss": 0.6976,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.003558525467704,
+      "learning_rate": 1.999607945235137e-05,
+      "loss": 0.8067,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8041757070189628,
+      "learning_rate": 1.9995986667742217e-05,
+      "loss": 0.7977,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5599959651379252,
+      "learning_rate": 1.999589279822503e-05,
+      "loss": 0.6672,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7795825546643322,
+      "learning_rate": 1.9995797843809998e-05,
+      "loss": 0.7974,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.802976071998248,
+      "learning_rate": 1.9995701804507423e-05,
+      "loss": 0.7719,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5941440848302881,
+      "learning_rate": 1.9995604680327735e-05,
+      "loss": 0.7913,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.753323779430216,
+      "learning_rate": 1.9995506471281473e-05,
+      "loss": 0.77,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6683717471946458,
+      "learning_rate": 1.9995407177379295e-05,
+      "loss": 0.8528,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6751255034212773,
+      "learning_rate": 1.9995306798631978e-05,
+      "loss": 0.7806,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7968947270822646,
+      "learning_rate": 1.999520533505042e-05,
+      "loss": 0.7733,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.025330454366504,
+      "learning_rate": 1.999510278664563e-05,
+      "loss": 0.7973,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5378582514197812,
+      "learning_rate": 1.9994999153428737e-05,
+      "loss": 0.6657,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.583531116731558,
+      "learning_rate": 1.999489443541099e-05,
+      "loss": 0.7282,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6735520561192077,
+      "learning_rate": 1.9994788632603756e-05,
+      "loss": 0.7124,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7029836297903527,
+      "learning_rate": 1.9994681745018516e-05,
+      "loss": 0.734,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.664386176676482,
+      "learning_rate": 1.999457377266687e-05,
+      "loss": 0.6538,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7530802960298322,
+      "learning_rate": 1.9994464715560543e-05,
+      "loss": 0.6985,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6860590312819281,
+      "learning_rate": 1.9994354573711363e-05,
+      "loss": 0.7206,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0901210609686722,
+      "learning_rate": 1.999424334713129e-05,
+      "loss": 0.7572,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8874014333140492,
+      "learning_rate": 1.9994131035832394e-05,
+      "loss": 0.782,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6553991557065093,
+      "learning_rate": 1.9994017639826867e-05,
+      "loss": 0.9557,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6718485788810615,
+      "learning_rate": 1.999390315912701e-05,
+      "loss": 0.7127,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7304276426396776,
+      "learning_rate": 1.9993787593745254e-05,
+      "loss": 0.7887,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8547453550237238,
+      "learning_rate": 1.999367094369414e-05,
+      "loss": 0.7698,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0608110132252917,
+      "learning_rate": 1.999355320898633e-05,
+      "loss": 0.7771,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6305518417499105,
+      "learning_rate": 1.99934343896346e-05,
+      "loss": 0.6427,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6217405497359352,
+      "learning_rate": 1.9993314485651848e-05,
+      "loss": 0.7834,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.620804556556247,
+      "learning_rate": 1.9993193497051084e-05,
+      "loss": 0.8264,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7909076263383104,
+      "learning_rate": 1.999307142384544e-05,
+      "loss": 0.8092,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.70064491469979,
+      "learning_rate": 1.9992948266048174e-05,
+      "loss": 0.821,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7059610993511671,
+      "learning_rate": 1.999282402367264e-05,
+      "loss": 0.7432,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5692469342874094,
+      "learning_rate": 1.999269869673233e-05,
+      "loss": 0.6727,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7378065054799623,
+      "learning_rate": 1.9992572285240848e-05,
+      "loss": 0.841,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8668129805207883,
+      "learning_rate": 1.999244478921191e-05,
+      "loss": 0.7503,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9060504553017656,
+      "learning_rate": 1.9992316208659352e-05,
+      "loss": 0.7054,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8212929976354297,
+      "learning_rate": 1.9992186543597133e-05,
+      "loss": 0.8107,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.809192974400173,
+      "learning_rate": 1.9992055794039325e-05,
+      "loss": 0.7067,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5465827774015835,
+      "learning_rate": 1.9991923960000116e-05,
+      "loss": 0.6662,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5850884512167587,
+      "learning_rate": 1.999179104149382e-05,
+      "loss": 0.7838,
+      "step": 418
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6622022535939016,
+      "learning_rate": 1.999165703853486e-05,
+      "loss": 0.7488,
+      "step": 419
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6903905078454566,
+      "learning_rate": 1.9991521951137783e-05,
+      "loss": 0.7443,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6687306292307453,
+      "learning_rate": 1.9991385779317246e-05,
+      "loss": 0.723,
+      "step": 421
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7219109927338176,
+      "learning_rate": 1.999124852308803e-05,
+      "loss": 0.78,
+      "step": 422
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8243785936412005,
+      "learning_rate": 1.9991110182465032e-05,
+      "loss": 0.6784,
+      "step": 423
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6490323132123654,
+      "learning_rate": 1.999097075746327e-05,
+      "loss": 0.6888,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7346088865508633,
+      "learning_rate": 1.999083024809787e-05,
+      "loss": 0.785,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6767037997897496,
+      "learning_rate": 1.999068865438409e-05,
+      "loss": 0.746,
+      "step": 426
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7701958205287336,
+      "learning_rate": 1.999054597633729e-05,
+      "loss": 0.7477,
+      "step": 427
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6835840347014255,
+      "learning_rate": 1.999040221397296e-05,
+      "loss": 0.7634,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7769492248134329,
+      "learning_rate": 1.99902573673067e-05,
+      "loss": 0.7342,
+      "step": 429
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7048105843168924,
+      "learning_rate": 1.9990111436354237e-05,
+      "loss": 0.6987,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5761305862194244,
+      "learning_rate": 1.99899644211314e-05,
+      "loss": 0.7776,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.74082648576072,
+      "learning_rate": 1.9989816321654155e-05,
+      "loss": 0.751,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8126076461111138,
+      "learning_rate": 1.9989667137938573e-05,
+      "loss": 0.7662,
+      "step": 433
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6477682572902752,
+      "learning_rate": 1.998951687000084e-05,
+      "loss": 0.7225,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7702013292110663,
+      "learning_rate": 1.998936551785727e-05,
+      "loss": 0.828,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7479554654572946,
+      "learning_rate": 1.9989213081524293e-05,
+      "loss": 0.7605,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7333751540111162,
+      "learning_rate": 1.9989059561018448e-05,
+      "loss": 0.7445,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8159978064450342,
+      "learning_rate": 1.9988904956356395e-05,
+      "loss": 0.7398,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6659874749639998,
+      "learning_rate": 1.998874926755492e-05,
+      "loss": 0.8073,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7995914872407046,
+      "learning_rate": 1.9988592494630922e-05,
+      "loss": 0.9021,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7213002946596805,
+      "learning_rate": 1.998843463760141e-05,
+      "loss": 0.7034,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7732611112379235,
+      "learning_rate": 1.998827569648352e-05,
+      "loss": 0.738,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0593757365062655,
+      "learning_rate": 1.9988115671294502e-05,
+      "loss": 0.8003,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5199583294558163,
+      "learning_rate": 1.9987954562051724e-05,
+      "loss": 0.7742,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6694614497357176,
+      "learning_rate": 1.9987792368772677e-05,
+      "loss": 0.7096,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.757071457535054,
+      "learning_rate": 1.9987629091474955e-05,
+      "loss": 0.7648,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4926165767089656,
+      "learning_rate": 1.9987464730176285e-05,
+      "loss": 0.6738,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6278470687032995,
+      "learning_rate": 1.9987299284894505e-05,
+      "loss": 0.6468,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5928480530517506,
+      "learning_rate": 1.9987132755647574e-05,
+      "loss": 0.733,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.954406246297614,
+      "learning_rate": 1.9986965142453562e-05,
+      "loss": 0.7253,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.677578009697896,
+      "learning_rate": 1.998679644533066e-05,
+      "loss": 0.7313,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.756632357254643,
+      "learning_rate": 1.998662666429718e-05,
+      "loss": 0.7917,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6989063531350133,
+      "learning_rate": 1.9986455799371555e-05,
+      "loss": 0.7377,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5734631073088516,
+      "learning_rate": 1.9986283850572317e-05,
+      "loss": 0.7689,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6438195775524247,
+      "learning_rate": 1.998611081791814e-05,
+      "loss": 0.6276,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7470830793024756,
+      "learning_rate": 1.9985936701427797e-05,
+      "loss": 0.8255,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.627502692382246,
+      "learning_rate": 1.998576150112019e-05,
+      "loss": 0.8686,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6742927725469114,
+      "learning_rate": 1.9985585217014326e-05,
+      "loss": 0.756,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.806975380520805,
+      "learning_rate": 1.9985407849129346e-05,
+      "loss": 0.8736,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6601591312990973,
+      "learning_rate": 1.9985229397484504e-05,
+      "loss": 0.7521,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.793242802316679,
+      "learning_rate": 1.9985049862099156e-05,
+      "loss": 0.7921,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6875996812673835,
+      "learning_rate": 1.9984869242992798e-05,
+      "loss": 0.8031,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5964038305874215,
+      "learning_rate": 1.9984687540185026e-05,
+      "loss": 0.63,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7710598946573337,
+      "learning_rate": 1.9984504753695567e-05,
+      "loss": 0.7269,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5836628328169005,
+      "learning_rate": 1.9984320883544254e-05,
+      "loss": 0.7293,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5838641530722382,
+      "learning_rate": 1.998413592975105e-05,
+      "loss": 0.6755,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4998685973317545,
+      "learning_rate": 1.9983949892336024e-05,
+      "loss": 0.6839,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8559871084679145,
+      "learning_rate": 1.998376277131937e-05,
+      "loss": 0.8468,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.794841563871742,
+      "learning_rate": 1.998357456672139e-05,
+      "loss": 0.7864,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6493411230580337,
+      "learning_rate": 1.9983385278562524e-05,
+      "loss": 0.7253,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8178196745533752,
+      "learning_rate": 1.9983194906863305e-05,
+      "loss": 0.8109,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6890060467396713,
+      "learning_rate": 1.99830034516444e-05,
+      "loss": 0.744,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7568201017705833,
+      "learning_rate": 1.9982810912926586e-05,
+      "loss": 0.8342,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6989251017754607,
+      "learning_rate": 1.998261729073076e-05,
+      "loss": 0.77,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.673235968940527,
+      "learning_rate": 1.998242258507794e-05,
+      "loss": 0.8354,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.777489272545835,
+      "learning_rate": 1.9982226795989255e-05,
+      "loss": 0.7293,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9007324938155739,
+      "learning_rate": 1.9982029923485957e-05,
+      "loss": 0.8945,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.695960720362699,
+      "learning_rate": 1.9981831967589412e-05,
+      "loss": 0.7604,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.546497291074308,
+      "learning_rate": 1.9981632928321103e-05,
+      "loss": 0.747,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7995388123021736,
+      "learning_rate": 1.9981432805702638e-05,
+      "loss": 0.7974,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8867834481521102,
+      "learning_rate": 1.998123159975573e-05,
+      "loss": 0.7991,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8366610120829818,
+      "learning_rate": 1.9981029310502227e-05,
+      "loss": 0.762,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5535863887360595,
+      "learning_rate": 1.9980825937964074e-05,
+      "loss": 0.8253,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6730007484155052,
+      "learning_rate": 1.9980621482163348e-05,
+      "loss": 0.6544,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.828936812673665,
+      "learning_rate": 1.998041594312224e-05,
+      "loss": 0.7863,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.669970570478486,
+      "learning_rate": 1.998020932086306e-05,
+      "loss": 0.8412,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.730258104372564,
+      "learning_rate": 1.9980001615408228e-05,
+      "loss": 0.8279,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7588302456093907,
+      "learning_rate": 1.997979282678029e-05,
+      "loss": 0.691,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5050103571137134,
+      "learning_rate": 1.9979582955001908e-05,
+      "loss": 0.6758,
+      "step": 489
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8806382255625589,
+      "learning_rate": 1.997937200009586e-05,
+      "loss": 0.7779,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7261954107457222,
+      "learning_rate": 1.9979159962085044e-05,
+      "loss": 0.7572,
+      "step": 491
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7225866766844808,
+      "learning_rate": 1.997894684099247e-05,
+      "loss": 0.8224,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6829420585393222,
+      "learning_rate": 1.9978732636841264e-05,
+      "loss": 0.6425,
+      "step": 493
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7874342491112851,
+      "learning_rate": 1.9978517349654684e-05,
+      "loss": 0.7899,
+      "step": 494
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.94124149740623,
+      "learning_rate": 1.9978300979456094e-05,
+      "loss": 0.7637,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.798068985507405,
+      "learning_rate": 1.9978083526268974e-05,
+      "loss": 0.9313,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7847619085647561,
+      "learning_rate": 1.9977864990116926e-05,
+      "loss": 0.8486,
+      "step": 497
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6565790972074341,
+      "learning_rate": 1.9977645371023673e-05,
+      "loss": 0.807,
+      "step": 498
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.654235902274819,
+      "learning_rate": 1.9977424669013044e-05,
+      "loss": 0.724,
+      "step": 499
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6431609076499223,
+      "learning_rate": 1.9977202884109e-05,
+      "loss": 0.6685,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.643346718045025,
+      "learning_rate": 1.9976980016335607e-05,
+      "loss": 0.7285,
+      "step": 501
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.833324399036458,
+      "learning_rate": 1.9976756065717056e-05,
+      "loss": 0.8175,
+      "step": 502
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6383228878544507,
+      "learning_rate": 1.9976531032277653e-05,
+      "loss": 0.736,
+      "step": 503
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9735369455703395,
+      "learning_rate": 1.9976304916041824e-05,
+      "loss": 0.6961,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7905953530647531,
+      "learning_rate": 1.9976077717034105e-05,
+      "loss": 0.8517,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.588535210734372,
+      "learning_rate": 1.997584943527916e-05,
+      "loss": 0.804,
+      "step": 506
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.627547386451281,
+      "learning_rate": 1.9975620070801763e-05,
+      "loss": 0.7761,
+      "step": 507
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7265931965712278,
+      "learning_rate": 1.997538962362681e-05,
+      "loss": 0.7825,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.847361586714884,
+      "learning_rate": 1.997515809377931e-05,
+      "loss": 0.7746,
+      "step": 509
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.759148489609264,
+      "learning_rate": 1.997492548128439e-05,
+      "loss": 0.8152,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.822969884000541,
+      "learning_rate": 1.9974691786167303e-05,
+      "loss": 0.7323,
+      "step": 511
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9328903420977737,
+      "learning_rate": 1.9974457008453408e-05,
+      "loss": 0.791,
+      "step": 512
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.687975853272304,
+      "learning_rate": 1.997422114816819e-05,
+      "loss": 0.7652,
+      "step": 513
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8072781375412006,
+      "learning_rate": 1.9973984205337244e-05,
+      "loss": 0.7159,
+      "step": 514
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7775120206705086,
+      "learning_rate": 1.997374617998629e-05,
+      "loss": 0.8137,
+      "step": 515
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6592282302457346,
+      "learning_rate": 1.997350707214116e-05,
+      "loss": 0.7775,
+      "step": 516
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6719635951730791,
+      "learning_rate": 1.99732668818278e-05,
+      "loss": 0.827,
+      "step": 517
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.65300911080359,
+      "learning_rate": 1.997302560907229e-05,
+      "loss": 0.7261,
+      "step": 518
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.76543409775746,
+      "learning_rate": 1.997278325390081e-05,
+      "loss": 0.8259,
+      "step": 519
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6797023411856506,
+      "learning_rate": 1.9972539816339664e-05,
+      "loss": 0.6834,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.816213182694269,
+      "learning_rate": 1.997229529641527e-05,
+      "loss": 0.8897,
+      "step": 521
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.583562441314376,
+      "learning_rate": 1.9972049694154175e-05,
+      "loss": 0.7364,
+      "step": 522
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5629931433036204,
+      "learning_rate": 1.997180300958303e-05,
+      "loss": 0.7036,
+      "step": 523
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7244295918749826,
+      "learning_rate": 1.997155524272861e-05,
+      "loss": 0.8148,
+      "step": 524
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9409193617437879,
+      "learning_rate": 1.997130639361781e-05,
+      "loss": 0.8473,
+      "step": 525
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7157934465546114,
+      "learning_rate": 1.9971056462277632e-05,
+      "loss": 0.7863,
+      "step": 526
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.646802444838358,
+      "learning_rate": 1.9970805448735205e-05,
+      "loss": 0.7295,
+      "step": 527
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.625090042380927,
+      "learning_rate": 1.9970553353017772e-05,
+      "loss": 0.7481,
+      "step": 528
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7295431401631154,
+      "learning_rate": 1.9970300175152696e-05,
+      "loss": 0.8117,
+      "step": 529
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6140197901048001,
+      "learning_rate": 1.997004591516745e-05,
+      "loss": 0.7754,
+      "step": 530
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6560281350322912,
+      "learning_rate": 1.9969790573089637e-05,
+      "loss": 0.6868,
+      "step": 531
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7529556342448862,
+      "learning_rate": 1.9969534148946965e-05,
+      "loss": 0.6683,
+      "step": 532
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7068734128772407,
+      "learning_rate": 1.996927664276727e-05,
+      "loss": 0.6886,
+      "step": 533
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.636282086916384,
+      "learning_rate": 1.9969018054578497e-05,
+      "loss": 0.707,
+      "step": 534
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.703668964082737,
+      "learning_rate": 1.9968758384408715e-05,
+      "loss": 0.8091,
+      "step": 535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6294150121655904,
+      "learning_rate": 1.99684976322861e-05,
+      "loss": 0.6922,
+      "step": 536
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7678562963178264,
+      "learning_rate": 1.9968235798238956e-05,
+      "loss": 0.8531,
+      "step": 537
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7785400771636022,
+      "learning_rate": 1.9967972882295704e-05,
+      "loss": 0.759,
+      "step": 538
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.028098506449123,
+      "learning_rate": 1.9967708884484875e-05,
+      "loss": 0.9118,
+      "step": 539
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5347148064228826,
+      "learning_rate": 1.996744380483513e-05,
+      "loss": 0.6544,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.820079444384609,
+      "learning_rate": 1.9967177643375227e-05,
+      "loss": 0.8159,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6671394749171597,
+      "learning_rate": 1.9966910400134062e-05,
+      "loss": 0.7221,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6788173634731034,
+      "learning_rate": 1.996664207514064e-05,
+      "loss": 0.7436,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5098704414985629,
+      "learning_rate": 1.9966372668424083e-05,
+      "loss": 0.6714,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5418486733686516,
+      "learning_rate": 1.9966102180013625e-05,
+      "loss": 0.758,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7685637436623223,
+      "learning_rate": 1.9965830609938627e-05,
+      "loss": 0.7747,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7274011670181888,
+      "learning_rate": 1.996555795822857e-05,
+      "loss": 0.8325,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6679061402459576,
+      "learning_rate": 1.9965284224913034e-05,
+      "loss": 0.8263,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.573807291130018,
+      "learning_rate": 1.9965009410021742e-05,
+      "loss": 0.713,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8231463328916808,
+      "learning_rate": 1.9964733513584508e-05,
+      "loss": 0.7005,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6741907583067228,
+      "learning_rate": 1.9964456535631287e-05,
+      "loss": 0.7334,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.79918584488658,
+      "learning_rate": 1.9964178476192132e-05,
+      "loss": 0.7676,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.811502444265952,
+      "learning_rate": 1.9963899335297227e-05,
+      "loss": 0.7321,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7755800902867478,
+      "learning_rate": 1.9963619112976867e-05,
+      "loss": 0.7925,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5597918230704464,
+      "learning_rate": 1.9963337809261465e-05,
+      "loss": 0.7272,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.864093123130176,
+      "learning_rate": 1.9963055424181556e-05,
+      "loss": 0.7669,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.389358668807301,
+      "learning_rate": 1.9962771957767782e-05,
+      "loss": 0.7292,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.74596830181998,
+      "learning_rate": 1.9962487410050915e-05,
+      "loss": 0.7526,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8385868687010367,
+      "learning_rate": 1.9962201781061833e-05,
+      "loss": 0.8057,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8668615928182508,
+      "learning_rate": 1.996191507083154e-05,
+      "loss": 0.792,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7652860661908676,
+      "learning_rate": 1.9961627279391154e-05,
+      "loss": 0.7552,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.739278918825244,
+      "learning_rate": 1.9961338406771908e-05,
+      "loss": 0.8162,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6244770953157561,
+      "learning_rate": 1.996104845300516e-05,
+      "loss": 0.7304,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7582338674779954,
+      "learning_rate": 1.996075741812237e-05,
+      "loss": 0.7494,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7029228717071425,
+      "learning_rate": 1.9960465302155132e-05,
+      "loss": 0.7869,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6435160351803817,
+      "learning_rate": 1.996017210513515e-05,
+      "loss": 0.7812,
+      "step": 566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6801997517535343,
+      "learning_rate": 1.9959877827094248e-05,
+      "loss": 0.765,
+      "step": 567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.835359427744583,
+      "learning_rate": 1.995958246806436e-05,
+      "loss": 0.8702,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7548856870112344,
+      "learning_rate": 1.995928602807755e-05,
+      "loss": 0.891,
+      "step": 569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6249171481144946,
+      "learning_rate": 1.9958988507165985e-05,
+      "loss": 0.7405,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8747074662765073,
+      "learning_rate": 1.9958689905361956e-05,
+      "loss": 0.8896,
+      "step": 571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.685467232191064,
+      "learning_rate": 1.9958390222697875e-05,
+      "loss": 0.8109,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5586476170652768,
+      "learning_rate": 1.9958089459206272e-05,
+      "loss": 0.8215,
+      "step": 573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9110312441730324,
+      "learning_rate": 1.9957787614919782e-05,
+      "loss": 0.683,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.795968523861863,
+      "learning_rate": 1.9957484689871167e-05,
+      "loss": 0.7298,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.619585532371068,
+      "learning_rate": 1.995718068409331e-05,
+      "loss": 0.8326,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.686148657322329,
+      "learning_rate": 1.99568755976192e-05,
+      "loss": 0.7904,
+      "step": 577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8980209427585717,
+      "learning_rate": 1.9956569430481954e-05,
+      "loss": 0.8522,
+      "step": 578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8223249537789892,
+      "learning_rate": 1.99562621827148e-05,
+      "loss": 0.7409,
+      "step": 579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6249663857855754,
+      "learning_rate": 1.9955953854351083e-05,
+      "loss": 0.7908,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6680154013641537,
+      "learning_rate": 1.995564444542427e-05,
+      "loss": 0.8155,
+      "step": 581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7437358002377406,
+      "learning_rate": 1.9955333955967936e-05,
+      "loss": 0.7572,
+      "step": 582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7442385392988993,
+      "learning_rate": 1.9955022386015792e-05,
+      "loss": 0.7381,
+      "step": 583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6455585855805404,
+      "learning_rate": 1.995470973560164e-05,
+      "loss": 0.7625,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5922361876502813,
+      "learning_rate": 1.995439600475943e-05,
+      "loss": 0.7293,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5827693035694907,
+      "learning_rate": 1.9954081193523197e-05,
+      "loss": 0.8177,
+      "step": 586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7369402384800399,
+      "learning_rate": 1.9953765301927116e-05,
+      "loss": 0.761,
+      "step": 587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7496504255649807,
+      "learning_rate": 1.9953448330005472e-05,
+      "loss": 0.7633,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6992475664068267,
+      "learning_rate": 1.9953130277792664e-05,
+      "loss": 0.7109,
+      "step": 589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7758089119133673,
+      "learning_rate": 1.9952811145323213e-05,
+      "loss": 0.8003,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5105006477745304,
+      "learning_rate": 1.995249093263176e-05,
+      "loss": 0.6333,
+      "step": 591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5506652898849551,
+      "learning_rate": 1.9952169639753055e-05,
+      "loss": 0.7387,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6406097273031046,
+      "learning_rate": 1.995184726672197e-05,
+      "loss": 0.754,
+      "step": 593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6937146704480206,
+      "learning_rate": 1.9951523813573496e-05,
+      "loss": 0.6996,
+      "step": 594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6863369020459251,
+      "learning_rate": 1.9951199280342732e-05,
+      "loss": 0.7959,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.601016536813647,
+      "learning_rate": 1.9950873667064906e-05,
+      "loss": 0.7021,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7116189222241849,
+      "learning_rate": 1.995054697377536e-05,
+      "loss": 0.7783,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7443096421429085,
+      "learning_rate": 1.9950219200509547e-05,
+      "loss": 0.6914,
+      "step": 598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8021785056836013,
+      "learning_rate": 1.9949890347303047e-05,
+      "loss": 0.7945,
+      "step": 599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6555910571792856,
+      "learning_rate": 1.9949560414191546e-05,
+      "loss": 0.7046,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7326941535085139,
+      "learning_rate": 1.9949229401210855e-05,
+      "loss": 0.8461,
+      "step": 601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6370922161012353,
+      "learning_rate": 1.9948897308396903e-05,
+      "loss": 0.6904,
+      "step": 602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6119600425769653,
+      "learning_rate": 1.9948564135785733e-05,
+      "loss": 0.8214,
+      "step": 603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7636336681905245,
+      "learning_rate": 1.9948229883413503e-05,
+      "loss": 0.8049,
+      "step": 604
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.766219115503329,
+      "learning_rate": 1.994789455131649e-05,
+      "loss": 0.6965,
+      "step": 605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7177562818571706,
+      "learning_rate": 1.994755813953109e-05,
+      "loss": 0.8001,
+      "step": 606
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.720211764446672,
+      "learning_rate": 1.994722064809382e-05,
+      "loss": 0.8151,
+      "step": 607
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7469479452575274,
+      "learning_rate": 1.9946882077041304e-05,
+      "loss": 0.7806,
+      "step": 608
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.85793386115385,
+      "learning_rate": 1.9946542426410295e-05,
+      "loss": 0.7966,
+      "step": 609
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7629818781691207,
+      "learning_rate": 1.9946201696237645e-05,
+      "loss": 0.7873,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6830394408562603,
+      "learning_rate": 1.9945859886560346e-05,
+      "loss": 0.7506,
+      "step": 611
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7860842728889232,
+      "learning_rate": 1.9945516997415493e-05,
+      "loss": 0.7918,
+      "step": 612
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6513821786445477,
+      "learning_rate": 1.9945173028840296e-05,
+      "loss": 0.8549,
+      "step": 613
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9218388455242161,
+      "learning_rate": 1.9944827980872094e-05,
+      "loss": 0.8566,
+      "step": 614
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7239023872462365,
+      "learning_rate": 1.9944481853548335e-05,
+      "loss": 0.7107,
+      "step": 615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7916304919841761,
+      "learning_rate": 1.9944134646906588e-05,
+      "loss": 0.8648,
+      "step": 616
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4579672832201422,
+      "learning_rate": 1.994378636098453e-05,
+      "loss": 0.7886,
+      "step": 617
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5409736030803547,
+      "learning_rate": 1.9943436995819968e-05,
+      "loss": 0.6151,
+      "step": 618
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7127356435439465,
+      "learning_rate": 1.9943086551450816e-05,
+      "loss": 0.7307,
+      "step": 619
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5792128695492997,
+      "learning_rate": 1.9942735027915113e-05,
+      "loss": 0.7638,
+      "step": 620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7555416285049137,
+      "learning_rate": 1.994238242525101e-05,
+      "loss": 0.8322,
+      "step": 621
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.805621761550607,
+      "learning_rate": 1.9942028743496773e-05,
+      "loss": 0.744,
+      "step": 622
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8466682699690558,
+      "learning_rate": 1.9941673982690795e-05,
+      "loss": 0.8617,
+      "step": 623
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6590526970977735,
+      "learning_rate": 1.9941318142871576e-05,
+      "loss": 0.6979,
+      "step": 624
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7786400594149645,
+      "learning_rate": 1.9940961224077736e-05,
+      "loss": 0.7248,
+      "step": 625
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6788965960428635,
+      "learning_rate": 1.9940603226348015e-05,
+      "loss": 0.7699,
+      "step": 626
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9265619143980974,
+      "learning_rate": 1.994024414972127e-05,
+      "loss": 0.8173,
+      "step": 627
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.715611931609657,
+      "learning_rate": 1.9939883994236468e-05,
+      "loss": 0.6624,
+      "step": 628
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5987849831467449,
+      "learning_rate": 1.99395227599327e-05,
+      "loss": 0.8136,
+      "step": 629
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6986730658985771,
+      "learning_rate": 1.9939160446849176e-05,
+      "loss": 0.7376,
+      "step": 630
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6545697033122497,
+      "learning_rate": 1.9938797055025213e-05,
+      "loss": 0.7149,
+      "step": 631
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.994617700256768,
+      "learning_rate": 1.993843258450026e-05,
+      "loss": 0.8161,
+      "step": 632
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8732615461651878,
+      "learning_rate": 1.9938067035313865e-05,
+      "loss": 0.7736,
+      "step": 633
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6800973995399204,
+      "learning_rate": 1.993770040750571e-05,
+      "loss": 0.8434,
+      "step": 634
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6925647661553942,
+      "learning_rate": 1.9937332701115585e-05,
+      "loss": 0.788,
+      "step": 635
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7217431415270548,
+      "learning_rate": 1.993696391618339e-05,
+      "loss": 0.6426,
+      "step": 636
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6743741825967342,
+      "learning_rate": 1.993659405274917e-05,
+      "loss": 0.7904,
+      "step": 637
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7108692982520572,
+      "learning_rate": 1.993622311085305e-05,
+      "loss": 0.6979,
+      "step": 638
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6256885715656448,
+      "learning_rate": 1.9935851090535295e-05,
+      "loss": 0.7019,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6229470491886617,
+      "learning_rate": 1.993547799183629e-05,
+      "loss": 0.7252,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5160911920409792,
+      "learning_rate": 1.9935103814796516e-05,
+      "loss": 0.7873,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5804423279608668,
+      "learning_rate": 1.9934728559456592e-05,
+      "loss": 0.8012,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4519598120061032,
+      "learning_rate": 1.9934352225857245e-05,
+      "loss": 0.7251,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7263353286073957,
+      "learning_rate": 1.993397481403932e-05,
+      "loss": 0.8427,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.637774815064037,
+      "learning_rate": 1.993359632404378e-05,
+      "loss": 0.7888,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7467785786117918,
+      "learning_rate": 1.9933216755911706e-05,
+      "loss": 0.7857,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5876144203148541,
+      "learning_rate": 1.9932836109684287e-05,
+      "loss": 0.7362,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.697451802836231,
+      "learning_rate": 1.993245438540284e-05,
+      "loss": 0.7894,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7166828340233518,
+      "learning_rate": 1.9932071583108796e-05,
+      "loss": 0.7418,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4878913435621886,
+      "learning_rate": 1.9931687702843706e-05,
+      "loss": 0.7117,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9177940612995616,
+      "learning_rate": 1.9931302744649224e-05,
+      "loss": 0.8025,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4139266215501305,
+      "learning_rate": 1.993091670856714e-05,
+      "loss": 0.7023,
+      "step": 652
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6439224416147458,
+      "learning_rate": 1.993052959463935e-05,
+      "loss": 0.7698,
+      "step": 653
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5886834224506095,
+      "learning_rate": 1.9930141402907867e-05,
+      "loss": 0.7759,
+      "step": 654
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5458636504703784,
+      "learning_rate": 1.9929752133414827e-05,
+      "loss": 0.7419,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5464792132429968,
+      "learning_rate": 1.9929361786202476e-05,
+      "loss": 0.6632,
+      "step": 656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.525133701113065,
+      "learning_rate": 1.9928970361313182e-05,
+      "loss": 0.8723,
+      "step": 657
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6124831659371417,
+      "learning_rate": 1.9928577858789424e-05,
+      "loss": 0.806,
+      "step": 658
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4701753543539475,
+      "learning_rate": 1.992818427867381e-05,
+      "loss": 0.6918,
+      "step": 659
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6849551194071541,
+      "learning_rate": 1.9927789621009045e-05,
+      "loss": 0.8426,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8884047795845,
+      "learning_rate": 1.9927393885837975e-05,
+      "loss": 0.695,
+      "step": 661
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8116947046420835,
+      "learning_rate": 1.9926997073203544e-05,
+      "loss": 0.8473,
+      "step": 662
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8689748277748406,
+      "learning_rate": 1.9926599183148822e-05,
+      "loss": 0.8169,
+      "step": 663
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.994878462441299,
+      "learning_rate": 1.9926200215716993e-05,
+      "loss": 0.8405,
+      "step": 664
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7987595227302102,
+      "learning_rate": 1.992580017095136e-05,
+      "loss": 0.7267,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5843737449052286,
+      "learning_rate": 1.992539904889534e-05,
+      "loss": 0.8127,
+      "step": 666
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8387810874585708,
+      "learning_rate": 1.992499684959247e-05,
+      "loss": 0.7091,
+      "step": 667
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7093730466761763,
+      "learning_rate": 1.9924593573086398e-05,
+      "loss": 0.7642,
+      "step": 668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7669829291305252,
+      "learning_rate": 1.99241892194209e-05,
+      "loss": 0.7177,
+      "step": 669
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7859363978876226,
+      "learning_rate": 1.9923783788639862e-05,
+      "loss": 0.7864,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5913637995967829,
+      "learning_rate": 1.992337728078728e-05,
+      "loss": 0.7669,
+      "step": 671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9006919289403583,
+      "learning_rate": 1.9922969695907278e-05,
+      "loss": 0.8227,
+      "step": 672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5707279920402772,
+      "learning_rate": 1.9922561034044095e-05,
+      "loss": 0.7132,
+      "step": 673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0511635474237364,
+      "learning_rate": 1.9922151295242085e-05,
+      "loss": 0.777,
+      "step": 674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.378420795935105,
+      "learning_rate": 1.9921740479545716e-05,
+      "loss": 0.6924,
+      "step": 675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3751596448402028,
+      "learning_rate": 1.9921328586999574e-05,
+      "loss": 0.7471,
+      "step": 676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6308793722944452,
+      "learning_rate": 1.992091561764837e-05,
+      "loss": 0.7539,
+      "step": 677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6200136862142986,
+      "learning_rate": 1.9920501571536917e-05,
+      "loss": 0.71,
+      "step": 678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6981092405574794,
+      "learning_rate": 1.9920086448710162e-05,
+      "loss": 0.7512,
+      "step": 679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.731471177175868,
+      "learning_rate": 1.9919670249213152e-05,
+      "loss": 0.8579,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6461346547255145,
+      "learning_rate": 1.9919252973091067e-05,
+      "loss": 0.7589,
+      "step": 681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7313768240322136,
+      "learning_rate": 1.991883462038919e-05,
+      "loss": 0.8429,
+      "step": 682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6033143635975582,
+      "learning_rate": 1.9918415191152927e-05,
+      "loss": 0.7439,
+      "step": 683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6283606197499367,
+      "learning_rate": 1.99179946854278e-05,
+      "loss": 0.7209,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6895382664512686,
+      "learning_rate": 1.9917573103259452e-05,
+      "loss": 0.8396,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5021274979284158,
+      "learning_rate": 1.9917150444693635e-05,
+      "loss": 0.809,
+      "step": 686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5898326262688758,
+      "learning_rate": 1.9916726709776228e-05,
+      "loss": 0.7391,
+      "step": 687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7143323205939232,
+      "learning_rate": 1.9916301898553215e-05,
+      "loss": 0.8389,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6601964607064166,
+      "learning_rate": 1.9915876011070705e-05,
+      "loss": 0.7384,
+      "step": 689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6997295899037483,
+      "learning_rate": 1.991544904737492e-05,
+      "loss": 0.6985,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6338386462513637,
+      "learning_rate": 1.9915021007512202e-05,
+      "loss": 0.8418,
+      "step": 691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6619689013467975,
+      "learning_rate": 1.991459189152901e-05,
+      "loss": 0.828,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5415233753313728,
+      "learning_rate": 1.991416169947191e-05,
+      "loss": 0.7842,
+      "step": 693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7594928368716753,
+      "learning_rate": 1.9913730431387603e-05,
+      "loss": 0.7791,
+      "step": 694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8285621513513672,
+      "learning_rate": 1.9913298087322886e-05,
+      "loss": 0.8216,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6189878588401247,
+      "learning_rate": 1.991286466732469e-05,
+      "loss": 0.7305,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6523217899208054,
+      "learning_rate": 1.9912430171440053e-05,
+      "loss": 0.8201,
+      "step": 697
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6392344267199934,
+      "learning_rate": 1.9911994599716137e-05,
+      "loss": 0.7456,
+      "step": 698
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.687768478994549,
+      "learning_rate": 1.9911557952200212e-05,
+      "loss": 0.7575,
+      "step": 699
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7800920776702511,
+      "learning_rate": 1.9911120228939668e-05,
+      "loss": 0.7296,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7376436911092747,
+      "learning_rate": 1.991068142998202e-05,
+      "loss": 0.8336,
+      "step": 701
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6366755441602803,
+      "learning_rate": 1.991024155537489e-05,
+      "loss": 0.6899,
+      "step": 702
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6311542633007308,
+      "learning_rate": 1.9909800605166013e-05,
+      "loss": 0.7963,
+      "step": 703
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6271019133596079,
+      "learning_rate": 1.990935857940325e-05,
+      "loss": 0.7225,
+      "step": 704
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.730023263842185,
+      "learning_rate": 1.9908915478134584e-05,
+      "loss": 0.8107,
+      "step": 705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7223462659250526,
+      "learning_rate": 1.9908471301408097e-05,
+      "loss": 0.8464,
+      "step": 706
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.766101862175367,
+      "learning_rate": 1.9908026049272006e-05,
+      "loss": 0.7489,
+      "step": 707
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5577791461925623,
+      "learning_rate": 1.990757972177463e-05,
+      "loss": 0.7676,
+      "step": 708
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7987244153202877,
+      "learning_rate": 1.990713231896441e-05,
+      "loss": 0.7746,
+      "step": 709
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7553912474460873,
+      "learning_rate": 1.990668384088991e-05,
+      "loss": 0.8975,
+      "step": 710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7169730984255511,
+      "learning_rate": 1.99062342875998e-05,
+      "loss": 0.7652,
+      "step": 711
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.826842707928925,
+      "learning_rate": 1.9905783659142878e-05,
+      "loss": 0.7199,
+      "step": 712
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7082093614480132,
+      "learning_rate": 1.9905331955568045e-05,
+      "loss": 0.694,
+      "step": 713
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5829316805497582,
+      "learning_rate": 1.990487917692433e-05,
+      "loss": 0.7717,
+      "step": 714
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6451440728127658,
+      "learning_rate": 1.9904425323260875e-05,
+      "loss": 0.7858,
+      "step": 715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.652820322856562,
+      "learning_rate": 1.990397039462694e-05,
+      "loss": 0.7959,
+      "step": 716
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7584740049172933,
+      "learning_rate": 1.99035143910719e-05,
+      "loss": 0.7561,
+      "step": 717
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7493478475508397,
+      "learning_rate": 1.990305731264525e-05,
+      "loss": 0.7176,
+      "step": 718
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.618316837166389,
+      "learning_rate": 1.990259915939659e-05,
+      "loss": 0.7485,
+      "step": 719
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6385372900613149,
+      "learning_rate": 1.9902139931375654e-05,
+      "loss": 0.8057,
+      "step": 720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5803614398859005,
+      "learning_rate": 1.9901679628632277e-05,
+      "loss": 0.6364,
+      "step": 721
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.620284724869963,
+      "learning_rate": 1.9901218251216424e-05,
+      "loss": 0.7402,
+      "step": 722
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.762040099161046,
+      "learning_rate": 1.9900755799178165e-05,
+      "loss": 0.7896,
+      "step": 723
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6331588276201974,
+      "learning_rate": 1.99002922725677e-05,
+      "loss": 0.7898,
+      "step": 724
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.650574167609646,
+      "learning_rate": 1.989982767143533e-05,
+      "loss": 0.7415,
+      "step": 725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8140348040677017,
+      "learning_rate": 1.9899361995831477e-05,
+      "loss": 0.9059,
+      "step": 726
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5401630519652945,
+      "learning_rate": 1.989889524580669e-05,
+      "loss": 0.6807,
+      "step": 727
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6906440729063186,
+      "learning_rate": 1.9898427421411627e-05,
+      "loss": 0.6651,
+      "step": 728
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7156359455773944,
+      "learning_rate": 1.9897958522697066e-05,
+      "loss": 0.7354,
+      "step": 729
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6778297011427266,
+      "learning_rate": 1.9897488549713887e-05,
+      "loss": 0.853,
+      "step": 730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5822425468702135,
+      "learning_rate": 1.9897017502513107e-05,
+      "loss": 0.8437,
+      "step": 731
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5697794417278896,
+      "learning_rate": 1.9896545381145854e-05,
+      "loss": 0.6893,
+      "step": 732
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6609595195224915,
+      "learning_rate": 1.9896072185663358e-05,
+      "loss": 0.8496,
+      "step": 733
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7210583504941446,
+      "learning_rate": 1.989559791611699e-05,
+      "loss": 0.7671,
+      "step": 734
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6351024582932636,
+      "learning_rate": 1.989512257255821e-05,
+      "loss": 0.7643,
+      "step": 735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5373211139173386,
+      "learning_rate": 1.9894646155038624e-05,
+      "loss": 0.7101,
+      "step": 736
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6909968283815922,
+      "learning_rate": 1.989416866360993e-05,
+      "loss": 0.7985,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5172315909593315,
+      "learning_rate": 1.9893690098323955e-05,
+      "loss": 0.6585,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6726015484769146,
+      "learning_rate": 1.989321045923264e-05,
+      "loss": 0.7843,
+      "step": 739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5799583071761338,
+      "learning_rate": 1.989272974638804e-05,
+      "loss": 0.8198,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6190253159018553,
+      "learning_rate": 1.9892247959842338e-05,
+      "loss": 0.7887,
+      "step": 741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6597510473945443,
+      "learning_rate": 1.989176509964781e-05,
+      "loss": 0.8144,
+      "step": 742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.535999847235126,
+      "learning_rate": 1.9891281165856876e-05,
+      "loss": 0.6152,
+      "step": 743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.766018263265675,
+      "learning_rate": 1.989079615852205e-05,
+      "loss": 0.7636,
+      "step": 744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5858962987014023,
+      "learning_rate": 1.9890310077695976e-05,
+      "loss": 0.7285,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5638476394295513,
+      "learning_rate": 1.988982292343141e-05,
+      "loss": 0.7238,
+      "step": 746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5665371791225642,
+      "learning_rate": 1.9889334695781227e-05,
+      "loss": 0.6451,
+      "step": 747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.629557149390182,
+      "learning_rate": 1.9888845394798416e-05,
+      "loss": 0.7883,
+      "step": 748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7545548160299196,
+      "learning_rate": 1.9888355020536078e-05,
+      "loss": 0.7631,
+      "step": 749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9006888455912045,
+      "learning_rate": 1.988786357304744e-05,
+      "loss": 0.7871,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7168503900092866,
+      "learning_rate": 1.9887371052385843e-05,
+      "loss": 0.8515,
+      "step": 751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8927518735252673,
+      "learning_rate": 1.9886877458604737e-05,
+      "loss": 0.8392,
+      "step": 752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7011730949471735,
+      "learning_rate": 1.9886382791757695e-05,
+      "loss": 0.7202,
+      "step": 753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.637575948589387,
+      "learning_rate": 1.9885887051898407e-05,
+      "loss": 0.74,
+      "step": 754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8131790572190998,
+      "learning_rate": 1.988539023908068e-05,
+      "loss": 0.7383,
+      "step": 755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6681631940285306,
+      "learning_rate": 1.988489235335843e-05,
+      "loss": 0.7128,
+      "step": 756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7593396966304202,
+      "learning_rate": 1.98843933947857e-05,
+      "loss": 0.8314,
+      "step": 757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7039767523301115,
+      "learning_rate": 1.988389336341664e-05,
+      "loss": 0.7352,
+      "step": 758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.679183477039409,
+      "learning_rate": 1.988339225930552e-05,
+      "loss": 0.617,
+      "step": 759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6711954196811725,
+      "learning_rate": 1.9882890082506733e-05,
+      "loss": 0.7709,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7192406625162853,
+      "learning_rate": 1.988238683307478e-05,
+      "loss": 0.9232,
+      "step": 761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5921764606177538,
+      "learning_rate": 1.9881882511064275e-05,
+      "loss": 0.7469,
+      "step": 762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6184206958606004,
+      "learning_rate": 1.9881377116529964e-05,
+      "loss": 0.681,
+      "step": 763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6011320408604426,
+      "learning_rate": 1.988087064952669e-05,
+      "loss": 0.7478,
+      "step": 764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4149338242204557,
+      "learning_rate": 1.9880363110109427e-05,
+      "loss": 0.7489,
+      "step": 765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6112484116223573,
+      "learning_rate": 1.987985449833326e-05,
+      "loss": 0.7996,
+      "step": 766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7167345566634746,
+      "learning_rate": 1.987934481425339e-05,
+      "loss": 0.8087,
+      "step": 767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5677451466214758,
+      "learning_rate": 1.987883405792514e-05,
+      "loss": 0.7465,
+      "step": 768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5804647890281336,
+      "learning_rate": 1.9878322229403938e-05,
+      "loss": 0.626,
+      "step": 769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.616554155661927,
+      "learning_rate": 1.9877809328745338e-05,
+      "loss": 0.6658,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5867376321137878,
+      "learning_rate": 1.987729535600501e-05,
+      "loss": 0.7254,
+      "step": 771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7469693174480554,
+      "learning_rate": 1.9876780311238727e-05,
+      "loss": 0.7281,
+      "step": 772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6938507016832554,
+      "learning_rate": 1.9876264194502403e-05,
+      "loss": 0.7879,
+      "step": 773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6085928749683316,
+      "learning_rate": 1.9875747005852048e-05,
+      "loss": 0.8294,
+      "step": 774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6144461068818,
+      "learning_rate": 1.987522874534379e-05,
+      "loss": 0.6786,
+      "step": 775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6544216920426305,
+      "learning_rate": 1.987470941303389e-05,
+      "loss": 0.7309,
+      "step": 776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7767428032177268,
+      "learning_rate": 1.9874189008978702e-05,
+      "loss": 0.7362,
+      "step": 777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.735644721679841,
+      "learning_rate": 1.9873667533234714e-05,
+      "loss": 0.8189,
+      "step": 778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7759896504317103,
+      "learning_rate": 1.987314498585852e-05,
+      "loss": 0.7809,
+      "step": 779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6443310875392836,
+      "learning_rate": 1.987262136690684e-05,
+      "loss": 0.7772,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6914137805454288,
+      "learning_rate": 1.9872096676436498e-05,
+      "loss": 0.6338,
+      "step": 781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9227888727294091,
+      "learning_rate": 1.9871570914504447e-05,
+      "loss": 0.8005,
+      "step": 782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.795882620074113,
+      "learning_rate": 1.9871044081167742e-05,
+      "loss": 0.7707,
+      "step": 783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6930358644809467,
+      "learning_rate": 1.9870516176483573e-05,
+      "loss": 0.8259,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7819340442518423,
+      "learning_rate": 1.9869987200509228e-05,
+      "loss": 0.6661,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.790035538160907,
+      "learning_rate": 1.9869457153302124e-05,
+      "loss": 0.829,
+      "step": 786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5202353379476825,
+      "learning_rate": 1.9868926034919787e-05,
+      "loss": 0.8845,
+      "step": 787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.863990678141754,
+      "learning_rate": 1.986839384541986e-05,
+      "loss": 0.7813,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5088174748801766,
+      "learning_rate": 1.9867860584860106e-05,
+      "loss": 0.8197,
+      "step": 789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.569596018469031,
+      "learning_rate": 1.98673262532984e-05,
+      "loss": 0.8397,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.76684943786916,
+      "learning_rate": 1.986679085079274e-05,
+      "loss": 0.8621,
+      "step": 791
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.909957245712091,
+      "learning_rate": 1.986625437740123e-05,
+      "loss": 0.9255,
+      "step": 792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.51112619591038,
+      "learning_rate": 1.98657168331821e-05,
+      "loss": 0.7438,
+      "step": 793
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7600103242211305,
+      "learning_rate": 1.986517821819369e-05,
+      "loss": 0.762,
+      "step": 794
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.629549944021964,
+      "learning_rate": 1.986463853249446e-05,
+      "loss": 0.8457,
+      "step": 795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6322779733282207,
+      "learning_rate": 1.9864097776142978e-05,
+      "loss": 0.7376,
+      "step": 796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6427720947747648,
+      "learning_rate": 1.9863555949197942e-05,
+      "loss": 0.7868,
+      "step": 797
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5070344335988768,
+      "learning_rate": 1.986301305171816e-05,
+      "loss": 0.7177,
+      "step": 798
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6517460173464724,
+      "learning_rate": 1.9862469083762546e-05,
+      "loss": 0.83,
+      "step": 799
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.862254713203288,
+      "learning_rate": 1.9861924045390147e-05,
+      "loss": 0.8391,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.840773968509843,
+      "learning_rate": 1.986137793666012e-05,
+      "loss": 0.7189,
+      "step": 801
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7091886103939837,
+      "learning_rate": 1.9860830757631727e-05,
+      "loss": 0.7196,
+      "step": 802
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.939228414732462,
+      "learning_rate": 1.9860282508364365e-05,
+      "loss": 0.7769,
+      "step": 803
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7527357282829756,
+      "learning_rate": 1.9859733188917532e-05,
+      "loss": 0.7744,
+      "step": 804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8700242525423052,
+      "learning_rate": 1.9859182799350852e-05,
+      "loss": 0.7007,
+      "step": 805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.595136671416076,
+      "learning_rate": 1.985863133972406e-05,
+      "loss": 0.7394,
+      "step": 806
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.666270275523058,
+      "learning_rate": 1.9858078810097004e-05,
+      "loss": 0.7128,
+      "step": 807
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6526420386804836,
+      "learning_rate": 1.985752521052966e-05,
+      "loss": 0.793,
+      "step": 808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0164674424468023,
+      "learning_rate": 1.9856970541082107e-05,
+      "loss": 0.7952,
+      "step": 809
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7422692497713812,
+      "learning_rate": 1.985641480181455e-05,
+      "loss": 0.7611,
+      "step": 810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8988225567936234,
+      "learning_rate": 1.98558579927873e-05,
+      "loss": 0.8972,
+      "step": 811
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5986381889416141,
+      "learning_rate": 1.9855300114060795e-05,
+      "loss": 0.7646,
+      "step": 812
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6483180495163923,
+      "learning_rate": 1.9854741165695583e-05,
+      "loss": 0.7311,
+      "step": 813
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8063581944455203,
+      "learning_rate": 1.9854181147752326e-05,
+      "loss": 0.9311,
+      "step": 814
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6504812770449069,
+      "learning_rate": 1.9853620060291812e-05,
+      "loss": 0.767,
+      "step": 815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4128816928019157,
+      "learning_rate": 1.985305790337493e-05,
+      "loss": 0.669,
+      "step": 816
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.733825753404519,
+      "learning_rate": 1.98524946770627e-05,
+      "loss": 0.8098,
+      "step": 817
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6299749868856428,
+      "learning_rate": 1.9851930381416243e-05,
+      "loss": 0.748,
+      "step": 818
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6517357971409825,
+      "learning_rate": 1.9851365016496812e-05,
+      "loss": 0.8209,
+      "step": 819
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.583252522074743,
+      "learning_rate": 1.985079858236577e-05,
+      "loss": 0.6843,
+      "step": 820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6536302219640018,
+      "learning_rate": 1.9850231079084593e-05,
+      "loss": 0.7518,
+      "step": 821
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7454823923954026,
+      "learning_rate": 1.9849662506714865e-05,
+      "loss": 0.7312,
+      "step": 822
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.914739489812267,
+      "learning_rate": 1.984909286531831e-05,
+      "loss": 0.8605,
+      "step": 823
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6639931309404836,
+      "learning_rate": 1.9848522154956744e-05,
+      "loss": 0.751,
+      "step": 824
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5651952261950397,
+      "learning_rate": 1.984795037569211e-05,
+      "loss": 0.7202,
+      "step": 825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5688629583006275,
+      "learning_rate": 1.984737752758647e-05,
+      "loss": 0.665,
+      "step": 826
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4482723619902158,
+      "learning_rate": 1.9846803610701996e-05,
+      "loss": 0.861,
+      "step": 827
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7229690482399078,
+      "learning_rate": 1.9846228625100977e-05,
+      "loss": 0.756,
+      "step": 828
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5180781551592952,
+      "learning_rate": 1.9845652570845818e-05,
+      "loss": 0.8151,
+      "step": 829
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.655586251600461,
+      "learning_rate": 1.9845075447999042e-05,
+      "loss": 0.8247,
+      "step": 830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6566059901082226,
+      "learning_rate": 1.9844497256623283e-05,
+      "loss": 0.7896,
+      "step": 831
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6097210602345795,
+      "learning_rate": 1.98439179967813e-05,
+      "loss": 0.7025,
+      "step": 832
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6245002946759295,
+      "learning_rate": 1.9843337668535958e-05,
+      "loss": 0.7641,
+      "step": 833
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5594374812846619,
+      "learning_rate": 1.9842756271950247e-05,
+      "loss": 0.6164,
+      "step": 834
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.662029547316825,
+      "learning_rate": 1.9842173807087264e-05,
+      "loss": 0.7241,
+      "step": 835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7204550225697404,
+      "learning_rate": 1.9841590274010228e-05,
+      "loss": 0.8256,
+      "step": 836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.712325268991625,
+      "learning_rate": 1.9841005672782473e-05,
+      "loss": 0.8094,
+      "step": 837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6558537995946654,
+      "learning_rate": 1.984042000346745e-05,
+      "loss": 0.8182,
+      "step": 838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4574707909678533,
+      "learning_rate": 1.9839833266128726e-05,
+      "loss": 0.7029,
+      "step": 839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5151941195461256,
+      "learning_rate": 1.9839245460829974e-05,
+      "loss": 0.8322,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.578641414260665,
+      "learning_rate": 1.9838656587634996e-05,
+      "loss": 0.6918,
+      "step": 841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.583223765478341,
+      "learning_rate": 1.983806664660771e-05,
+      "loss": 0.6128,
+      "step": 842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.35378598549046,
+      "learning_rate": 1.9837475637812132e-05,
+      "loss": 0.6721,
+      "step": 843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5124064601223453,
+      "learning_rate": 1.983688356131242e-05,
+      "loss": 0.7138,
+      "step": 844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6007621242000454,
+      "learning_rate": 1.9836290417172825e-05,
+      "loss": 0.7293,
+      "step": 845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.532861934209684,
+      "learning_rate": 1.983569620545773e-05,
+      "loss": 0.7525,
+      "step": 846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5919170282152117,
+      "learning_rate": 1.9835100926231625e-05,
+      "loss": 0.7039,
+      "step": 847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.613310000105187,
+      "learning_rate": 1.9834504579559118e-05,
+      "loss": 0.7526,
+      "step": 848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5091381434192657,
+      "learning_rate": 1.9833907165504935e-05,
+      "loss": 0.674,
+      "step": 849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6288949471182157,
+      "learning_rate": 1.9833308684133913e-05,
+      "loss": 0.6881,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7035249813071198,
+      "learning_rate": 1.9832709135511006e-05,
+      "loss": 0.7646,
+      "step": 851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6085841109244907,
+      "learning_rate": 1.983210851970129e-05,
+      "loss": 0.7064,
+      "step": 852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5416377825068288,
+      "learning_rate": 1.9831506836769955e-05,
+      "loss": 0.6925,
+      "step": 853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6407129976757333,
+      "learning_rate": 1.9830904086782298e-05,
+      "loss": 0.8426,
+      "step": 854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.711869418845939,
+      "learning_rate": 1.983030026980374e-05,
+      "loss": 0.8236,
+      "step": 855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.511105480844578,
+      "learning_rate": 1.9829695385899816e-05,
+      "loss": 0.7293,
+      "step": 856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.727960597652724,
+      "learning_rate": 1.9829089435136176e-05,
+      "loss": 0.7806,
+      "step": 857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8756242422118425,
+      "learning_rate": 1.982848241757859e-05,
+      "loss": 0.8235,
+      "step": 858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.751065062669428,
+      "learning_rate": 1.982787433329294e-05,
+      "loss": 0.8057,
+      "step": 859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5458333978687038,
+      "learning_rate": 1.9827265182345218e-05,
+      "loss": 0.7308,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5226977001977733,
+      "learning_rate": 1.9826654964801544e-05,
+      "loss": 0.7669,
+      "step": 861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6839225185773306,
+      "learning_rate": 1.9826043680728142e-05,
+      "loss": 0.8051,
+      "step": 862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5793830832920863,
+      "learning_rate": 1.9825431330191366e-05,
+      "loss": 0.8132,
+      "step": 863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6937724188454104,
+      "learning_rate": 1.9824817913257666e-05,
+      "loss": 0.6887,
+      "step": 864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.873772989542665,
+      "learning_rate": 1.9824203429993627e-05,
+      "loss": 0.8393,
+      "step": 865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6477348786228707,
+      "learning_rate": 1.982358788046594e-05,
+      "loss": 0.8163,
+      "step": 866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.52524445412708,
+      "learning_rate": 1.9822971264741412e-05,
+      "loss": 0.7488,
+      "step": 867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6276327048556951,
+      "learning_rate": 1.9822353582886963e-05,
+      "loss": 0.6686,
+      "step": 868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.634620053938659,
+      "learning_rate": 1.9821734834969643e-05,
+      "loss": 0.8702,
+      "step": 869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5984651127834377,
+      "learning_rate": 1.98211150210566e-05,
+      "loss": 0.6718,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5710084200353578,
+      "learning_rate": 1.98204941412151e-05,
+      "loss": 0.6835,
+      "step": 871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6624153524943914,
+      "learning_rate": 1.9819872195512545e-05,
+      "loss": 0.7358,
+      "step": 872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.590680428823555,
+      "learning_rate": 1.9819249184016426e-05,
+      "loss": 0.7424,
+      "step": 873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6679560046276667,
+      "learning_rate": 1.9818625106794363e-05,
+      "loss": 0.7768,
+      "step": 874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.576313274431883,
+      "learning_rate": 1.981799996391409e-05,
+      "loss": 0.758,
+      "step": 875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6518281038340517,
+      "learning_rate": 1.981737375544346e-05,
+      "loss": 0.7233,
+      "step": 876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4274151645512967,
+      "learning_rate": 1.9816746481450436e-05,
+      "loss": 0.7168,
+      "step": 877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.939294035861291,
+      "learning_rate": 1.9816118142003096e-05,
+      "loss": 0.8503,
+      "step": 878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5387673753810336,
+      "learning_rate": 1.981548873716964e-05,
+      "loss": 0.7362,
+      "step": 879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4805195280344556,
+      "learning_rate": 1.9814858267018376e-05,
+      "loss": 0.6436,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5446391199954803,
+      "learning_rate": 1.9814226731617734e-05,
+      "loss": 0.6923,
+      "step": 881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4813312189994743,
+      "learning_rate": 1.981359413103626e-05,
+      "loss": 0.7571,
+      "step": 882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6640726456178598,
+      "learning_rate": 1.981296046534261e-05,
+      "loss": 0.851,
+      "step": 883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6633157436957888,
+      "learning_rate": 1.981232573460556e-05,
+      "loss": 0.7917,
+      "step": 884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6746676607899385,
+      "learning_rate": 1.9811689938893998e-05,
+      "loss": 0.7473,
+      "step": 885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7031648068997371,
+      "learning_rate": 1.9811053078276933e-05,
+      "loss": 0.7654,
+      "step": 886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5023261382337536,
+      "learning_rate": 1.981041515282348e-05,
+      "loss": 0.6898,
+      "step": 887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5128621826976616,
+      "learning_rate": 1.980977616260288e-05,
+      "loss": 0.6342,
+      "step": 888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7107190654157496,
+      "learning_rate": 1.980913610768449e-05,
+      "loss": 0.8767,
+      "step": 889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7018440255203298,
+      "learning_rate": 1.9808494988137766e-05,
+      "loss": 0.8087,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.748811212451861,
+      "learning_rate": 1.9807852804032306e-05,
+      "loss": 0.7345,
+      "step": 891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7272712489265187,
+      "learning_rate": 1.98072095554378e-05,
+      "loss": 0.8743,
+      "step": 892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8069453963982343,
+      "learning_rate": 1.9806565242424064e-05,
+      "loss": 0.8003,
+      "step": 893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5389870763989857,
+      "learning_rate": 1.980591986506103e-05,
+      "loss": 0.7899,
+      "step": 894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3762396601862672,
+      "learning_rate": 1.9805273423418737e-05,
+      "loss": 0.7024,
+      "step": 895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.46629201856903,
+      "learning_rate": 1.980462591756735e-05,
+      "loss": 0.7164,
+      "step": 896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6408767957620132,
+      "learning_rate": 1.9803977347577154e-05,
+      "loss": 0.7415,
+      "step": 897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5881132334026928,
+      "learning_rate": 1.980332771351853e-05,
+      "loss": 0.8025,
+      "step": 898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5763669524638544,
+      "learning_rate": 1.980267701546199e-05,
+      "loss": 0.8883,
+      "step": 899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4978780824228621,
+      "learning_rate": 1.980202525347816e-05,
+      "loss": 0.7626,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4887098145165154,
+      "learning_rate": 1.980137242763777e-05,
+      "loss": 0.8167,
+      "step": 901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7711545128982842,
+      "learning_rate": 1.9800718538011683e-05,
+      "loss": 0.7537,
+      "step": 902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5460885788878576,
+      "learning_rate": 1.9800063584670864e-05,
+      "loss": 0.7343,
+      "step": 903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7386866230563274,
+      "learning_rate": 1.97994075676864e-05,
+      "loss": 0.8492,
+      "step": 904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5807749512434859,
+      "learning_rate": 1.9798750487129486e-05,
+      "loss": 0.6922,
+      "step": 905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5324171436351097,
+      "learning_rate": 1.9798092343071445e-05,
+      "loss": 0.7796,
+      "step": 906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7562281371712432,
+      "learning_rate": 1.9797433135583705e-05,
+      "loss": 0.7693,
+      "step": 907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6217028713794182,
+      "learning_rate": 1.979677286473781e-05,
+      "loss": 0.7588,
+      "step": 908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.565090135593504,
+      "learning_rate": 1.9796111530605428e-05,
+      "loss": 0.7376,
+      "step": 909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.603068989358705,
+      "learning_rate": 1.9795449133258335e-05,
+      "loss": 0.7798,
+      "step": 910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6396418489523752,
+      "learning_rate": 1.979478567276842e-05,
+      "loss": 0.6676,
+      "step": 911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6372104385022896,
+      "learning_rate": 1.979412114920769e-05,
+      "loss": 0.7513,
+      "step": 912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.675396810421105,
+      "learning_rate": 1.9793455562648276e-05,
+      "loss": 0.7631,
+      "step": 913
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.897849675524275,
+      "learning_rate": 1.9792788913162414e-05,
+      "loss": 0.7721,
+      "step": 914
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5554235700494816,
+      "learning_rate": 1.9792121200822456e-05,
+      "loss": 0.81,
+      "step": 915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4373647032594903,
+      "learning_rate": 1.9791452425700874e-05,
+      "loss": 0.5976,
+      "step": 916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7491234619263865,
+      "learning_rate": 1.9790782587870252e-05,
+      "loss": 0.7904,
+      "step": 917
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6486451154873125,
+      "learning_rate": 1.9790111687403292e-05,
+      "loss": 0.7869,
+      "step": 918
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5310571807878468,
+      "learning_rate": 1.978943972437281e-05,
+      "loss": 0.7273,
+      "step": 919
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7047828597241863,
+      "learning_rate": 1.978876669885173e-05,
+      "loss": 0.682,
+      "step": 920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6301078272480116,
+      "learning_rate": 1.9788092610913107e-05,
+      "loss": 0.799,
+      "step": 921
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5601134386377202,
+      "learning_rate": 1.97874174606301e-05,
+      "loss": 0.8432,
+      "step": 922
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7210324057395259,
+      "learning_rate": 1.9786741248075984e-05,
+      "loss": 0.8726,
+      "step": 923
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.394576608403927,
+      "learning_rate": 1.9786063973324156e-05,
+      "loss": 0.7333,
+      "step": 924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5976279760603629,
+      "learning_rate": 1.978538563644812e-05,
+      "loss": 0.7835,
+      "step": 925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5422919699984112,
+      "learning_rate": 1.97847062375215e-05,
+      "loss": 0.772,
+      "step": 926
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6557365404347686,
+      "learning_rate": 1.978402577661803e-05,
+      "loss": 0.7868,
+      "step": 927
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.674019352909184,
+      "learning_rate": 1.978334425381157e-05,
+      "loss": 0.7464,
+      "step": 928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6255566180377932,
+      "learning_rate": 1.9782661669176086e-05,
+      "loss": 0.7936,
+      "step": 929
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6899360192659847,
+      "learning_rate": 1.978197802278566e-05,
+      "loss": 0.8837,
+      "step": 930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5427340894121977,
+      "learning_rate": 1.9781293314714493e-05,
+      "loss": 0.7157,
+      "step": 931
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5486328001346643,
+      "learning_rate": 1.97806075450369e-05,
+      "loss": 0.6634,
+      "step": 932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.676171028632494,
+      "learning_rate": 1.9779920713827307e-05,
+      "loss": 0.6629,
+      "step": 933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6169928978186605,
+      "learning_rate": 1.9779232821160264e-05,
+      "loss": 0.7855,
+      "step": 934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6601303576342932,
+      "learning_rate": 1.9778543867110428e-05,
+      "loss": 0.7458,
+      "step": 935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.572233190447626,
+      "learning_rate": 1.9777853851752575e-05,
+      "loss": 0.7343,
+      "step": 936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5786433225350176,
+      "learning_rate": 1.9777162775161592e-05,
+      "loss": 0.7364,
+      "step": 937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6711676113294462,
+      "learning_rate": 1.977647063741249e-05,
+      "loss": 0.6002,
+      "step": 938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6719456821034266,
+      "learning_rate": 1.9775777438580387e-05,
+      "loss": 0.7643,
+      "step": 939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6821802539166844,
+      "learning_rate": 1.977508317874052e-05,
+      "loss": 0.8292,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5626244212813953,
+      "learning_rate": 1.9774387857968237e-05,
+      "loss": 0.6857,
+      "step": 941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.621394503878368,
+      "learning_rate": 1.977369147633901e-05,
+      "loss": 0.6925,
+      "step": 942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5671624337676704,
+      "learning_rate": 1.977299403392841e-05,
+      "loss": 0.7112,
+      "step": 943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.004929948433543,
+      "learning_rate": 1.9772295530812143e-05,
+      "loss": 0.7504,
+      "step": 944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5805032467684188,
+      "learning_rate": 1.977159596706602e-05,
+      "loss": 0.7705,
+      "step": 945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.564866358149519,
+      "learning_rate": 1.9770895342765964e-05,
+      "loss": 0.6859,
+      "step": 946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5104736001360075,
+      "learning_rate": 1.977019365798802e-05,
+      "loss": 0.6725,
+      "step": 947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7279061948601635,
+      "learning_rate": 1.976949091280834e-05,
+      "loss": 0.8203,
+      "step": 948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6987389794038148,
+      "learning_rate": 1.97687871073032e-05,
+      "loss": 0.7356,
+      "step": 949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5683557461074862,
+      "learning_rate": 1.976808224154899e-05,
+      "loss": 0.7154,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6291551489710554,
+      "learning_rate": 1.9767376315622204e-05,
+      "loss": 0.6979,
+      "step": 951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7429785759185206,
+      "learning_rate": 1.9766669329599466e-05,
+      "loss": 0.9519,
+      "step": 952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7201420018698663,
+      "learning_rate": 1.9765961283557503e-05,
+      "loss": 0.8157,
+      "step": 953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6340952247661684,
+      "learning_rate": 1.976525217757317e-05,
+      "loss": 0.8053,
+      "step": 954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7497449502423785,
+      "learning_rate": 1.976454201172342e-05,
+      "loss": 0.6892,
+      "step": 955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8490925913199021,
+      "learning_rate": 1.976383078608534e-05,
+      "loss": 0.7282,
+      "step": 956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.571142660205738,
+      "learning_rate": 1.976311850073611e-05,
+      "loss": 0.7532,
+      "step": 957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.490756284704002,
+      "learning_rate": 1.976240515575305e-05,
+      "loss": 0.7327,
+      "step": 958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6843788399658988,
+      "learning_rate": 1.976169075121358e-05,
+      "loss": 0.7708,
+      "step": 959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7748135897626862,
+      "learning_rate": 1.976097528719523e-05,
+      "loss": 0.8446,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6707936681278435,
+      "learning_rate": 1.9760258763775656e-05,
+      "loss": 0.7498,
+      "step": 961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7033717941830635,
+      "learning_rate": 1.975954118103263e-05,
+      "loss": 0.6827,
+      "step": 962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6250184779189945,
+      "learning_rate": 1.975882253904403e-05,
+      "loss": 0.8026,
+      "step": 963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6361132028766527,
+      "learning_rate": 1.9758102837887853e-05,
+      "loss": 0.6545,
+      "step": 964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6965149086218916,
+      "learning_rate": 1.9757382077642214e-05,
+      "loss": 0.8264,
+      "step": 965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6633583574231603,
+      "learning_rate": 1.9756660258385338e-05,
+      "loss": 0.8511,
+      "step": 966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5977541239908342,
+      "learning_rate": 1.975593738019557e-05,
+      "loss": 0.7447,
+      "step": 967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.56727108125653,
+      "learning_rate": 1.975521344315136e-05,
+      "loss": 0.7605,
+      "step": 968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.622962864888057,
+      "learning_rate": 1.9754488447331292e-05,
+      "loss": 0.7268,
+      "step": 969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5541156726666745,
+      "learning_rate": 1.9753762392814043e-05,
+      "loss": 0.7134,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.610775363823247,
+      "learning_rate": 1.975303527967842e-05,
+      "loss": 0.7554,
+      "step": 971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7499914820423736,
+      "learning_rate": 1.9752307108003334e-05,
+      "loss": 0.8552,
+      "step": 972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5534873957880497,
+      "learning_rate": 1.9751577877867823e-05,
+      "loss": 0.753,
+      "step": 973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.808776873744801,
+      "learning_rate": 1.975084758935103e-05,
+      "loss": 0.7819,
+      "step": 974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7700864769108424,
+      "learning_rate": 1.9750116242532217e-05,
+      "loss": 0.9487,
+      "step": 975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6943439977730848,
+      "learning_rate": 1.974938383749076e-05,
+      "loss": 0.7968,
+      "step": 976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6407491012547737,
+      "learning_rate": 1.9748650374306156e-05,
+      "loss": 0.7705,
+      "step": 977
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.70469858711636,
+      "learning_rate": 1.9747915853058003e-05,
+      "loss": 0.7693,
+      "step": 978
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.582707324059266,
+      "learning_rate": 1.974718027382602e-05,
+      "loss": 0.7874,
+      "step": 979
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6073999705030813,
+      "learning_rate": 1.974644363669005e-05,
+      "loss": 0.799,
+      "step": 980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6479941375707847,
+      "learning_rate": 1.974570594173004e-05,
+      "loss": 0.7785,
+      "step": 981
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6646530813302085,
+      "learning_rate": 1.974496718902606e-05,
+      "loss": 0.8484,
+      "step": 982
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.506444754411822,
+      "learning_rate": 1.9744227378658283e-05,
+      "loss": 0.7749,
+      "step": 983
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6978448175283514,
+      "learning_rate": 1.9743486510707006e-05,
+      "loss": 0.9159,
+      "step": 984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.542376548702729,
+      "learning_rate": 1.974274458525264e-05,
+      "loss": 0.7079,
+      "step": 985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6363243613427414,
+      "learning_rate": 1.9742001602375708e-05,
+      "loss": 0.802,
+      "step": 986
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.596726103662305,
+      "learning_rate": 1.9741257562156854e-05,
+      "loss": 0.8283,
+      "step": 987
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.552072210203679,
+      "learning_rate": 1.974051246467682e-05,
+      "loss": 0.7828,
+      "step": 988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7601591894326951,
+      "learning_rate": 1.973976631001649e-05,
+      "loss": 0.6845,
+      "step": 989
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.751601699761324,
+      "learning_rate": 1.9739019098256835e-05,
+      "loss": 0.7943,
+      "step": 990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5746857081818262,
+      "learning_rate": 1.973827082947896e-05,
+      "loss": 0.7584,
+      "step": 991
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6786168546308895,
+      "learning_rate": 1.9737521503764076e-05,
+      "loss": 0.704,
+      "step": 992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7689656599100716,
+      "learning_rate": 1.9736771121193513e-05,
+      "loss": 0.8223,
+      "step": 993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6628613066396887,
+      "learning_rate": 1.9736019681848706e-05,
+      "loss": 0.7639,
+      "step": 994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6507086963589646,
+      "learning_rate": 1.973526718581122e-05,
+      "loss": 0.7448,
+      "step": 995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4891511625875535,
+      "learning_rate": 1.9734513633162723e-05,
+      "loss": 0.7129,
+      "step": 996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.801981972115585,
+      "learning_rate": 1.9733759023985e-05,
+      "loss": 0.7863,
+      "step": 997
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6483362524488605,
+      "learning_rate": 1.9733003358359955e-05,
+      "loss": 0.6735,
+      "step": 998
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5831381134582354,
+      "learning_rate": 1.9732246636369605e-05,
+      "loss": 0.74,
+      "step": 999
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7761677747689284,
+      "learning_rate": 1.9731488858096078e-05,
+      "loss": 0.7121,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5907097418790517,
+      "learning_rate": 1.9730730023621617e-05,
+      "loss": 0.7825,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6103489754656453,
+      "learning_rate": 1.972997013302858e-05,
+      "loss": 0.8179,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5474773072539885,
+      "learning_rate": 1.9729209186399448e-05,
+      "loss": 0.7446,
+      "step": 1003
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.522546952331723,
+      "learning_rate": 1.972844718381681e-05,
+      "loss": 0.7528,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.78490021522784,
+      "learning_rate": 1.9727684125363364e-05,
+      "loss": 0.887,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5860785797184678,
+      "learning_rate": 1.972692001112193e-05,
+      "loss": 0.71,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5648733823815915,
+      "learning_rate": 1.972615484117544e-05,
+      "loss": 0.7583,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.707739890292607,
+      "learning_rate": 1.972538861560694e-05,
+      "loss": 0.7862,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.659828837091845,
+      "learning_rate": 1.9724621334499597e-05,
+      "loss": 0.7663,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6810106139429293,
+      "learning_rate": 1.9723852997936683e-05,
+      "loss": 0.7745,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5375832455600957,
+      "learning_rate": 1.972308360600159e-05,
+      "loss": 0.7442,
+      "step": 1011
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8275172545359188,
+      "learning_rate": 1.9722313158777825e-05,
+      "loss": 0.8308,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5558075737762853,
+      "learning_rate": 1.9721541656349005e-05,
+      "loss": 0.7765,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7634963149801517,
+      "learning_rate": 1.9720769098798866e-05,
+      "loss": 0.7983,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5644790225602312,
+      "learning_rate": 1.9719995486211258e-05,
+      "loss": 0.7695,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5350140296556833,
+      "learning_rate": 1.9719220818670144e-05,
+      "loss": 0.8209,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.716896406028085,
+      "learning_rate": 1.97184450962596e-05,
+      "loss": 0.8061,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5783961891694578,
+      "learning_rate": 1.971766831906382e-05,
+      "loss": 0.6355,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8176065622672664,
+      "learning_rate": 1.9716890487167114e-05,
+      "loss": 0.7767,
+      "step": 1019
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5373269041490218,
+      "learning_rate": 1.97161116006539e-05,
+      "loss": 0.8069,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.779558739717626,
+      "learning_rate": 1.9715331659608716e-05,
+      "loss": 0.9195,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5333585600073345,
+      "learning_rate": 1.971455066411621e-05,
+      "loss": 0.762,
+      "step": 1022
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.568751380846371,
+      "learning_rate": 1.9713768614261144e-05,
+      "loss": 0.7352,
+      "step": 1023
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4626406800636604,
+      "learning_rate": 1.9712985510128406e-05,
+      "loss": 0.7571,
+      "step": 1024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6115763573627133,
+      "learning_rate": 1.9712201351802985e-05,
+      "loss": 0.8721,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4664746057207612,
+      "learning_rate": 1.9711416139369984e-05,
+      "loss": 0.6534,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.834484523255519,
+      "learning_rate": 1.971062987291464e-05,
+      "loss": 0.7143,
+      "step": 1027
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5306636799352586,
+      "learning_rate": 1.9709842552522272e-05,
+      "loss": 0.7261,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.480494949221711,
+      "learning_rate": 1.9709054178278343e-05,
+      "loss": 0.7775,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4159653760073085,
+      "learning_rate": 1.9708264750268416e-05,
+      "loss": 0.6398,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4677707044356991,
+      "learning_rate": 1.9707474268578172e-05,
+      "loss": 0.8,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6534741622675229,
+      "learning_rate": 1.97066827332934e-05,
+      "loss": 0.802,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5389456821966196,
+      "learning_rate": 1.9705890144500012e-05,
+      "loss": 0.764,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.58608861866648,
+      "learning_rate": 1.9705096502284037e-05,
+      "loss": 0.6967,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7745355167995958,
+      "learning_rate": 1.9704301806731604e-05,
+      "loss": 0.7554,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6537554040171483,
+      "learning_rate": 1.970350605792897e-05,
+      "loss": 0.7989,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.49864081600279,
+      "learning_rate": 1.97027092559625e-05,
+      "loss": 0.6163,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6541958709909896,
+      "learning_rate": 1.9701911400918673e-05,
+      "loss": 0.7903,
+      "step": 1038
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7003266196100684,
+      "learning_rate": 1.9701112492884084e-05,
+      "loss": 0.8901,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6807657449170552,
+      "learning_rate": 1.9700312531945444e-05,
+      "loss": 0.6864,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.614703409405987,
+      "learning_rate": 1.969951151818957e-05,
+      "loss": 0.839,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7806689503190112,
+      "learning_rate": 1.9698709451703405e-05,
+      "loss": 0.7984,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8040362335997895,
+      "learning_rate": 1.9697906332574005e-05,
+      "loss": 0.8109,
+      "step": 1043
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7385609277311922,
+      "learning_rate": 1.9697102160888525e-05,
+      "loss": 0.7235,
+      "step": 1044
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6972535125809556,
+      "learning_rate": 1.9696296936734254e-05,
+      "loss": 0.7874,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8012082058922911,
+      "learning_rate": 1.9695490660198584e-05,
+      "loss": 0.7251,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7649284575263353,
+      "learning_rate": 1.9694683331369023e-05,
+      "loss": 0.7602,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6764194604703575,
+      "learning_rate": 1.9693874950333196e-05,
+      "loss": 0.8882,
+      "step": 1048
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6350072171486167,
+      "learning_rate": 1.9693065517178836e-05,
+      "loss": 0.6515,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4989978096130447,
+      "learning_rate": 1.9692255031993794e-05,
+      "loss": 0.7783,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3890840164822351,
+      "learning_rate": 1.9691443494866043e-05,
+      "loss": 0.5872,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.63102381931439,
+      "learning_rate": 1.9690630905883658e-05,
+      "loss": 0.7515,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6734613888623586,
+      "learning_rate": 1.968981726513483e-05,
+      "loss": 0.7333,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.643404966679851,
+      "learning_rate": 1.9689002572707873e-05,
+      "loss": 0.7412,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.601773214091372,
+      "learning_rate": 1.96881868286912e-05,
+      "loss": 0.7213,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6130619273018145,
+      "learning_rate": 1.9687370033173356e-05,
+      "loss": 0.7485,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7587685900576053,
+      "learning_rate": 1.968655218624299e-05,
+      "loss": 0.767,
+      "step": 1057
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6537766624915502,
+      "learning_rate": 1.968573328798886e-05,
+      "loss": 0.6481,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7185152830818846,
+      "learning_rate": 1.9684913338499855e-05,
+      "loss": 0.8093,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9319626256997462,
+      "learning_rate": 1.968409233786496e-05,
+      "loss": 0.6413,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.547975529846523,
+      "learning_rate": 1.9683270286173286e-05,
+      "loss": 0.749,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6573271155926854,
+      "learning_rate": 1.968244718351405e-05,
+      "loss": 0.644,
+      "step": 1062
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5989053921550134,
+      "learning_rate": 1.968162302997659e-05,
+      "loss": 0.818,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7268069657969647,
+      "learning_rate": 1.968079782565035e-05,
+      "loss": 0.7992,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.862673950648877,
+      "learning_rate": 1.9679971570624902e-05,
+      "loss": 0.7751,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6890329705049392,
+      "learning_rate": 1.9679144264989914e-05,
+      "loss": 0.7416,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6130513551926382,
+      "learning_rate": 1.967831590883518e-05,
+      "loss": 0.7199,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.653095405272077,
+      "learning_rate": 1.967748650225061e-05,
+      "loss": 0.7696,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.445353967928603,
+      "learning_rate": 1.9676656045326217e-05,
+      "loss": 0.7718,
+      "step": 1069
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6023578260531115,
+      "learning_rate": 1.967582453815214e-05,
+      "loss": 0.8398,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6055120340482538,
+      "learning_rate": 1.967499198081862e-05,
+      "loss": 0.8376,
+      "step": 1071
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5238960023480306,
+      "learning_rate": 1.9674158373416017e-05,
+      "loss": 0.619,
+      "step": 1072
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7324188075148437,
+      "learning_rate": 1.9673323716034812e-05,
+      "loss": 0.9763,
+      "step": 1073
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5746141870115677,
+      "learning_rate": 1.967248800876559e-05,
+      "loss": 0.7339,
+      "step": 1074
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4853638045612656,
+      "learning_rate": 1.967165125169906e-05,
+      "loss": 0.7768,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7272333668866946,
+      "learning_rate": 1.9670813444926028e-05,
+      "loss": 0.8803,
+      "step": 1076
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7776062101599435,
+      "learning_rate": 1.9669974588537437e-05,
+      "loss": 0.7372,
+      "step": 1077
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7740405473527927,
+      "learning_rate": 1.9669134682624324e-05,
+      "loss": 0.7725,
+      "step": 1078
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6327544110262224,
+      "learning_rate": 1.9668293727277847e-05,
+      "loss": 0.8023,
+      "step": 1079
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6830258720590778,
+      "learning_rate": 1.9667451722589283e-05,
+      "loss": 0.8351,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.632865983465764,
+      "learning_rate": 1.966660866865002e-05,
+      "loss": 0.8502,
+      "step": 1081
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8501165148542782,
+      "learning_rate": 1.9665764565551548e-05,
+      "loss": 0.8346,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.643792627111494,
+      "learning_rate": 1.966491941338549e-05,
+      "loss": 0.6841,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7869989049285269,
+      "learning_rate": 1.9664073212243576e-05,
+      "loss": 0.7794,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5272749364519826,
+      "learning_rate": 1.966322596221764e-05,
+      "loss": 0.6827,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.548034982280869,
+      "learning_rate": 1.9662377663399647e-05,
+      "loss": 0.7078,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7707301221752616,
+      "learning_rate": 1.9661528315881654e-05,
+      "loss": 0.8737,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.738629441499874,
+      "learning_rate": 1.9660677919755855e-05,
+      "loss": 0.9194,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6275322704385589,
+      "learning_rate": 1.9659826475114548e-05,
+      "loss": 0.7833,
+      "step": 1089
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6804171948269853,
+      "learning_rate": 1.9658973982050133e-05,
+      "loss": 0.776,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.575697569245944,
+      "learning_rate": 1.9658120440655147e-05,
+      "loss": 0.7907,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.570696801563854,
+      "learning_rate": 1.965726585102222e-05,
+      "loss": 0.7582,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6602049450510845,
+      "learning_rate": 1.965641021324411e-05,
+      "loss": 0.718,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6935903967630492,
+      "learning_rate": 1.965555352741368e-05,
+      "loss": 0.8315,
+      "step": 1094
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.795982914953094,
+      "learning_rate": 1.9654695793623908e-05,
+      "loss": 0.7485,
+      "step": 1095
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5328332724743226,
+      "learning_rate": 1.9653837011967894e-05,
+      "loss": 0.7178,
+      "step": 1096
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5640940025238694,
+      "learning_rate": 1.9652977182538836e-05,
+      "loss": 0.7228,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.556543305513743,
+      "learning_rate": 1.9652116305430067e-05,
+      "loss": 0.8408,
+      "step": 1098
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5417627207071147,
+      "learning_rate": 1.965125438073501e-05,
+      "loss": 0.7884,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.526098291521299,
+      "learning_rate": 1.9650391408547222e-05,
+      "loss": 0.8267,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7507073343126842,
+      "learning_rate": 1.9649527388960365e-05,
+      "loss": 0.891,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5063836112608697,
+      "learning_rate": 1.9648662322068205e-05,
+      "loss": 0.8313,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6884409011943324,
+      "learning_rate": 1.964779620796464e-05,
+      "loss": 0.7544,
+      "step": 1103
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7060068858942647,
+      "learning_rate": 1.9646929046743675e-05,
+      "loss": 0.727,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3902547802147753,
+      "learning_rate": 1.9646060838499418e-05,
+      "loss": 0.7339,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4974045504752067,
+      "learning_rate": 1.964519158332611e-05,
+      "loss": 0.8947,
+      "step": 1106
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.627185223593396,
+      "learning_rate": 1.9644321281318085e-05,
+      "loss": 0.7223,
+      "step": 1107
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8650431831681953,
+      "learning_rate": 1.9643449932569803e-05,
+      "loss": 0.8473,
+      "step": 1108
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5295919963150972,
+      "learning_rate": 1.9642577537175845e-05,
+      "loss": 0.6756,
+      "step": 1109
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.827817218480238,
+      "learning_rate": 1.9641704095230884e-05,
+      "loss": 0.7595,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8572345935296157,
+      "learning_rate": 1.9640829606829724e-05,
+      "loss": 0.8726,
+      "step": 1111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9229900375324138,
+      "learning_rate": 1.9639954072067273e-05,
+      "loss": 0.6821,
+      "step": 1112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.529709612228805,
+      "learning_rate": 1.9639077491038562e-05,
+      "loss": 0.7728,
+      "step": 1113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7816911587916253,
+      "learning_rate": 1.9638199863838726e-05,
+      "loss": 0.6983,
+      "step": 1114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5904056602001073,
+      "learning_rate": 1.9637321190563023e-05,
+      "loss": 0.7543,
+      "step": 1115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.764043464464645,
+      "learning_rate": 1.963644147130681e-05,
+      "loss": 0.8764,
+      "step": 1116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.581725970823919,
+      "learning_rate": 1.9635560706165577e-05,
+      "loss": 0.7722,
+      "step": 1117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.483170971872246,
+      "learning_rate": 1.9634678895234912e-05,
+      "loss": 0.6994,
+      "step": 1118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.607073912269842,
+      "learning_rate": 1.963379603861052e-05,
+      "loss": 0.7012,
+      "step": 1119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5788678057053491,
+      "learning_rate": 1.9632912136388228e-05,
+      "loss": 0.6462,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.736077450846907,
+      "learning_rate": 1.9632027188663963e-05,
+      "loss": 0.7634,
+      "step": 1121
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.553299674016011,
+      "learning_rate": 1.9631141195533773e-05,
+      "loss": 0.7819,
+      "step": 1122
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4890833928925393,
+      "learning_rate": 1.9630254157093822e-05,
+      "loss": 0.6707,
+      "step": 1123
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7022152845461826,
+      "learning_rate": 1.9629366073440385e-05,
+      "loss": 0.7831,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.574794997754779,
+      "learning_rate": 1.9628476944669847e-05,
+      "loss": 0.6835,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6830359629247464,
+      "learning_rate": 1.9627586770878705e-05,
+      "loss": 0.7132,
+      "step": 1126
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6489366566650268,
+      "learning_rate": 1.962669555216358e-05,
+      "loss": 0.8291,
+      "step": 1127
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5845939269600287,
+      "learning_rate": 1.9625803288621194e-05,
+      "loss": 0.7917,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4958457966941456,
+      "learning_rate": 1.96249099803484e-05,
+      "loss": 0.7724,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6052748434750355,
+      "learning_rate": 1.9624015627442134e-05,
+      "loss": 0.7444,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5071294567920583,
+      "learning_rate": 1.962312022999948e-05,
+      "loss": 0.6666,
+      "step": 1131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7281784144036083,
+      "learning_rate": 1.9622223788117614e-05,
+      "loss": 0.7376,
+      "step": 1132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6185475613897287,
+      "learning_rate": 1.962132630189383e-05,
+      "loss": 0.7193,
+      "step": 1133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7260084083877627,
+      "learning_rate": 1.962042777142554e-05,
+      "loss": 0.7864,
+      "step": 1134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5682532904630293,
+      "learning_rate": 1.9619528196810254e-05,
+      "loss": 0.7225,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6391159674374254,
+      "learning_rate": 1.9618627578145624e-05,
+      "loss": 0.7729,
+      "step": 1136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5784690928165177,
+      "learning_rate": 1.9617725915529383e-05,
+      "loss": 0.8187,
+      "step": 1137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7088910412836227,
+      "learning_rate": 1.9616823209059398e-05,
+      "loss": 0.784,
+      "step": 1138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6571345846430185,
+      "learning_rate": 1.9615919458833646e-05,
+      "loss": 0.6829,
+      "step": 1139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6319202664587704,
+      "learning_rate": 1.9615014664950214e-05,
+      "loss": 0.75,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4303103285508485,
+      "learning_rate": 1.9614108827507304e-05,
+      "loss": 0.6818,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5585376987437913,
+      "learning_rate": 1.9613201946603225e-05,
+      "loss": 0.8776,
+      "step": 1142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6595202034800935,
+      "learning_rate": 1.9612294022336407e-05,
+      "loss": 0.8057,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8047651958531572,
+      "learning_rate": 1.9611385054805396e-05,
+      "loss": 0.8138,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5737996224499786,
+      "learning_rate": 1.9610475044108837e-05,
+      "loss": 0.6541,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5709415443250103,
+      "learning_rate": 1.960956399034551e-05,
+      "loss": 0.7679,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.58783726351291,
+      "learning_rate": 1.9608651893614284e-05,
+      "loss": 0.772,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6528514722614565,
+      "learning_rate": 1.960773875401416e-05,
+      "loss": 0.6852,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6218521782436417,
+      "learning_rate": 1.960682457164424e-05,
+      "loss": 0.7883,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4396576356085227,
+      "learning_rate": 1.960590934660375e-05,
+      "loss": 0.6858,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6058391146974962,
+      "learning_rate": 1.9604993078992016e-05,
+      "loss": 0.6883,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.752370009104151,
+      "learning_rate": 1.960407576890849e-05,
+      "loss": 0.8184,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4172074465364604,
+      "learning_rate": 1.9603157416452732e-05,
+      "loss": 0.6789,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.843728995407731,
+      "learning_rate": 1.960223802172441e-05,
+      "loss": 0.7007,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.631743660794764,
+      "learning_rate": 1.9601317584823315e-05,
+      "loss": 0.764,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.484817707471324,
+      "learning_rate": 1.9600396105849346e-05,
+      "loss": 0.7767,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6070133386997218,
+      "learning_rate": 1.9599473584902512e-05,
+      "loss": 0.8539,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.695249667390617,
+      "learning_rate": 1.9598550022082937e-05,
+      "loss": 0.7938,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6234052698473143,
+      "learning_rate": 1.9597625417490863e-05,
+      "loss": 0.757,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6293595346347398,
+      "learning_rate": 1.9596699771226638e-05,
+      "loss": 0.7461,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.662704230882139,
+      "learning_rate": 1.959577308339073e-05,
+      "loss": 0.926,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.756056604825178,
+      "learning_rate": 1.9594845354083716e-05,
+      "loss": 0.8179,
+      "step": 1162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6719212295165158,
+      "learning_rate": 1.9593916583406285e-05,
+      "loss": 0.7861,
+      "step": 1163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6184719245446453,
+      "learning_rate": 1.959298677145924e-05,
+      "loss": 0.8487,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7206162783405348,
+      "learning_rate": 1.95920559183435e-05,
+      "loss": 0.7311,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8359997424317043,
+      "learning_rate": 1.9591124024160092e-05,
+      "loss": 0.7866,
+      "step": 1166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.624456920621825,
+      "learning_rate": 1.959019108901016e-05,
+      "loss": 0.747,
+      "step": 1167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6260806359564066,
+      "learning_rate": 1.9589257112994955e-05,
+      "loss": 0.69,
+      "step": 1168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.667073327058875,
+      "learning_rate": 1.958832209621586e-05,
+      "loss": 0.7738,
+      "step": 1169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6441841038248202,
+      "learning_rate": 1.9587386038774336e-05,
+      "loss": 0.7193,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.621706152411868,
+      "learning_rate": 1.958644894077199e-05,
+      "loss": 0.8084,
+      "step": 1171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.586482320038094,
+      "learning_rate": 1.9585510802310527e-05,
+      "loss": 0.7456,
+      "step": 1172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.692881629428406,
+      "learning_rate": 1.9584571623491767e-05,
+      "loss": 0.8248,
+      "step": 1173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7294424016547736,
+      "learning_rate": 1.9583631404417644e-05,
+      "loss": 0.7689,
+      "step": 1174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6193899899241604,
+      "learning_rate": 1.9582690145190203e-05,
+      "loss": 0.8272,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7799529369465272,
+      "learning_rate": 1.9581747845911606e-05,
+      "loss": 0.8271,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.493448273391503,
+      "learning_rate": 1.958080450668412e-05,
+      "loss": 0.6895,
+      "step": 1177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5369375142395612,
+      "learning_rate": 1.9579860127610127e-05,
+      "loss": 0.7262,
+      "step": 1178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5421982874414422,
+      "learning_rate": 1.9578914708792137e-05,
+      "loss": 0.7941,
+      "step": 1179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.595212877359824,
+      "learning_rate": 1.9577968250332748e-05,
+      "loss": 0.8563,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.55371549416413,
+      "learning_rate": 1.957702075233469e-05,
+      "loss": 0.7626,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4194941949887323,
+      "learning_rate": 1.9576072214900797e-05,
+      "loss": 0.783,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5571594964699573,
+      "learning_rate": 1.9575122638134018e-05,
+      "loss": 0.6418,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6443867895825042,
+      "learning_rate": 1.9574172022137416e-05,
+      "loss": 0.7109,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5964204320215944,
+      "learning_rate": 1.9573220367014164e-05,
+      "loss": 0.6518,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.593349684628786,
+      "learning_rate": 1.9572267672867546e-05,
+      "loss": 0.7596,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.616352916585672,
+      "learning_rate": 1.957131393980097e-05,
+      "loss": 0.6699,
+      "step": 1187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.623342101148859,
+      "learning_rate": 1.9570359167917942e-05,
+      "loss": 0.753,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5982626120798151,
+      "learning_rate": 1.956940335732209e-05,
+      "loss": 0.774,
+      "step": 1189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5780506501342173,
+      "learning_rate": 1.9568446508117155e-05,
+      "loss": 0.6698,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6191888496783715,
+      "learning_rate": 1.9567488620406984e-05,
+      "loss": 0.7069,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.128084240737797,
+      "learning_rate": 1.956652969429554e-05,
+      "loss": 0.8099,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.654400131567404,
+      "learning_rate": 1.95655697298869e-05,
+      "loss": 0.8374,
+      "step": 1193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5711157493370986,
+      "learning_rate": 1.9564608727285258e-05,
+      "loss": 0.6901,
+      "step": 1194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.458281055944847,
+      "learning_rate": 1.956364668659491e-05,
+      "loss": 0.6599,
+      "step": 1195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6305506241103176,
+      "learning_rate": 1.9562683607920278e-05,
+      "loss": 0.8076,
+      "step": 1196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6685254009668504,
+      "learning_rate": 1.956171949136588e-05,
+      "loss": 0.7673,
+      "step": 1197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5926444627846572,
+      "learning_rate": 1.956075433703636e-05,
+      "loss": 0.6434,
+      "step": 1198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5334041797983586,
+      "learning_rate": 1.955978814503647e-05,
+      "loss": 0.6648,
+      "step": 1199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6348793124114884,
+      "learning_rate": 1.9558820915471076e-05,
+      "loss": 0.8082,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6650518092286242,
+      "learning_rate": 1.9557852648445155e-05,
+      "loss": 0.7144,
+      "step": 1201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7730364636444929,
+      "learning_rate": 1.95568833440638e-05,
+      "loss": 0.8038,
+      "step": 1202
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4817291628359017,
+      "learning_rate": 1.9555913002432207e-05,
+      "loss": 0.7215,
+      "step": 1203
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5437673000618928,
+      "learning_rate": 1.9554941623655695e-05,
+      "loss": 0.7794,
+      "step": 1204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4958564172666966,
+      "learning_rate": 1.9553969207839694e-05,
+      "loss": 0.7292,
+      "step": 1205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4671382704484248,
+      "learning_rate": 1.955299575508974e-05,
+      "loss": 0.718,
+      "step": 1206
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5760005500160656,
+      "learning_rate": 1.9552021265511492e-05,
+      "loss": 0.7143,
+      "step": 1207
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5823479034445453,
+      "learning_rate": 1.955104573921071e-05,
+      "loss": 0.7174,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7896892822695007,
+      "learning_rate": 1.955006917629328e-05,
+      "loss": 0.8204,
+      "step": 1209
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.609040548686237,
+      "learning_rate": 1.9549091576865186e-05,
+      "loss": 0.7619,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5210302834438783,
+      "learning_rate": 1.954811294103253e-05,
+      "loss": 0.6917,
+      "step": 1211
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.668606437182086,
+      "learning_rate": 1.9547133268901527e-05,
+      "loss": 0.7154,
+      "step": 1212
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7072837999942514,
+      "learning_rate": 1.954615256057851e-05,
+      "loss": 0.7634,
+      "step": 1213
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8353300138895317,
+      "learning_rate": 1.954517081616992e-05,
+      "loss": 0.7178,
+      "step": 1214
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5972113260052339,
+      "learning_rate": 1.954418803578231e-05,
+      "loss": 0.7412,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6412684612277957,
+      "learning_rate": 1.9543204219522338e-05,
+      "loss": 0.728,
+      "step": 1216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.609735472945734,
+      "learning_rate": 1.954221936749679e-05,
+      "loss": 0.8719,
+      "step": 1217
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5573591820655903,
+      "learning_rate": 1.9541233479812552e-05,
+      "loss": 0.7166,
+      "step": 1218
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7781813815209297,
+      "learning_rate": 1.9540246556576624e-05,
+      "loss": 0.7628,
+      "step": 1219
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.683038497796432,
+      "learning_rate": 1.9539258597896132e-05,
+      "loss": 0.9519,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6477867242414523,
+      "learning_rate": 1.9538269603878293e-05,
+      "loss": 0.7761,
+      "step": 1221
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6247177428236637,
+      "learning_rate": 1.953727957463045e-05,
+      "loss": 0.8392,
+      "step": 1222
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7641582896967056,
+      "learning_rate": 1.953628851026006e-05,
+      "loss": 0.8251,
+      "step": 1223
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.651347180107763,
+      "learning_rate": 1.9535296410874678e-05,
+      "loss": 0.7091,
+      "step": 1224
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.596795753990509,
+      "learning_rate": 1.953430327658199e-05,
+      "loss": 0.6776,
+      "step": 1225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.72023293687916,
+      "learning_rate": 1.9533309107489775e-05,
+      "loss": 0.7708,
+      "step": 1226
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8641043621843565,
+      "learning_rate": 1.9532313903705945e-05,
+      "loss": 0.8081,
+      "step": 1227
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6892750119679534,
+      "learning_rate": 1.953131766533851e-05,
+      "loss": 0.8025,
+      "step": 1228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6979908514604627,
+      "learning_rate": 1.9530320392495592e-05,
+      "loss": 0.8839,
+      "step": 1229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5614345333542738,
+      "learning_rate": 1.9529322085285436e-05,
+      "loss": 0.8074,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5462404519794781,
+      "learning_rate": 1.952832274381639e-05,
+      "loss": 0.7008,
+      "step": 1231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6801337840039725,
+      "learning_rate": 1.9527322368196913e-05,
+      "loss": 0.7842,
+      "step": 1232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4877774460670745,
+      "learning_rate": 1.952632095853559e-05,
+      "loss": 0.7126,
+      "step": 1233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.470108974836532,
+      "learning_rate": 1.9525318514941097e-05,
+      "loss": 0.7865,
+      "step": 1234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6233458319969059,
+      "learning_rate": 1.9524315037522238e-05,
+      "loss": 0.7215,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6854954491546525,
+      "learning_rate": 1.9523310526387928e-05,
+      "loss": 0.7766,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.71560710244287,
+      "learning_rate": 1.9522304981647186e-05,
+      "loss": 0.8284,
+      "step": 1237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5165468607991677,
+      "learning_rate": 1.952129840340915e-05,
+      "loss": 0.7682,
+      "step": 1238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5311283978080743,
+      "learning_rate": 1.952029079178307e-05,
+      "loss": 0.7429,
+      "step": 1239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5454786126577658,
+      "learning_rate": 1.951928214687831e-05,
+      "loss": 0.7592,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7566208659424567,
+      "learning_rate": 1.9518272468804333e-05,
+      "loss": 0.7453,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6901545265480546,
+      "learning_rate": 1.9517261757670728e-05,
+      "loss": 0.8635,
+      "step": 1242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7770640309026147,
+      "learning_rate": 1.9516250013587193e-05,
+      "loss": 0.7617,
+      "step": 1243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5424774446781067,
+      "learning_rate": 1.9515237236663537e-05,
+      "loss": 0.7539,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6504202538179722,
+      "learning_rate": 1.951422342700968e-05,
+      "loss": 0.7253,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.479830693335698,
+      "learning_rate": 1.9513208584735656e-05,
+      "loss": 0.7184,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.610053631904774,
+      "learning_rate": 1.9512192709951613e-05,
+      "loss": 0.8248,
+      "step": 1247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.647736990225513,
+      "learning_rate": 1.95111758027678e-05,
+      "loss": 0.7153,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6762468882847545,
+      "learning_rate": 1.9510157863294595e-05,
+      "loss": 0.8218,
+      "step": 1249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5976749648522812,
+      "learning_rate": 1.9509138891642476e-05,
+      "loss": 0.7339,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6550024682871,
+      "learning_rate": 1.950811888792204e-05,
+      "loss": 0.754,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8202097251988651,
+      "learning_rate": 1.9507097852243982e-05,
+      "loss": 0.7897,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5527467626965263,
+      "learning_rate": 1.950607578471913e-05,
+      "loss": 0.7427,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.431786500229432,
+      "learning_rate": 1.950505268545841e-05,
+      "loss": 0.6891,
+      "step": 1254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5255211241113107,
+      "learning_rate": 1.9504028554572865e-05,
+      "loss": 0.6895,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.573716102171307,
+      "learning_rate": 1.9503003392173643e-05,
+      "loss": 0.7392,
+      "step": 1256
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7500915942835833,
+      "learning_rate": 1.9501977198372017e-05,
+      "loss": 0.8491,
+      "step": 1257
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6546658273474368,
+      "learning_rate": 1.9500949973279358e-05,
+      "loss": 0.6988,
+      "step": 1258
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6133269018317669,
+      "learning_rate": 1.949992171700716e-05,
+      "loss": 0.6442,
+      "step": 1259
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6354850724656633,
+      "learning_rate": 1.949889242966702e-05,
+      "loss": 0.8259,
+      "step": 1260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7682967900115991,
+      "learning_rate": 1.9497862111370654e-05,
+      "loss": 0.8345,
+      "step": 1261
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6029947100999593,
+      "learning_rate": 1.9496830762229884e-05,
+      "loss": 0.8014,
+      "step": 1262
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6075492800959517,
+      "learning_rate": 1.949579838235665e-05,
+      "loss": 0.7311,
+      "step": 1263
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6998815244012917,
+      "learning_rate": 1.9494764971862998e-05,
+      "loss": 0.7245,
+      "step": 1264
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4951211694261486,
+      "learning_rate": 1.949373053086109e-05,
+      "loss": 0.8232,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5067111018633088,
+      "learning_rate": 1.9492695059463197e-05,
+      "loss": 0.7533,
+      "step": 1266
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6037628206146815,
+      "learning_rate": 1.949165855778171e-05,
+      "loss": 0.7363,
+      "step": 1267
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5768105452349963,
+      "learning_rate": 1.9490621025929112e-05,
+      "loss": 0.7573,
+      "step": 1268
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.01346038116964,
+      "learning_rate": 1.9489582464018023e-05,
+      "loss": 0.8341,
+      "step": 1269
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7239135570187674,
+      "learning_rate": 1.948854287216116e-05,
+      "loss": 0.7889,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4433082279559541,
+      "learning_rate": 1.9487502250471347e-05,
+      "loss": 0.6745,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6967479347098455,
+      "learning_rate": 1.9486460599061536e-05,
+      "loss": 0.6947,
+      "step": 1272
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6368410087098855,
+      "learning_rate": 1.948541791804478e-05,
+      "loss": 0.7861,
+      "step": 1273
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7263202574572991,
+      "learning_rate": 1.948437420753424e-05,
+      "loss": 0.7657,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4448531385811203,
+      "learning_rate": 1.94833294676432e-05,
+      "loss": 0.5998,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5124154039368294,
+      "learning_rate": 1.948228369848505e-05,
+      "loss": 0.7219,
+      "step": 1276
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.702412187559942,
+      "learning_rate": 1.948123690017329e-05,
+      "loss": 0.781,
+      "step": 1277
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.555786535491567,
+      "learning_rate": 1.9480189072821537e-05,
+      "loss": 0.7099,
+      "step": 1278
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6032400807672151,
+      "learning_rate": 1.947914021654351e-05,
+      "loss": 0.6508,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3784201916828709,
+      "learning_rate": 1.9478090331453054e-05,
+      "loss": 0.7026,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5784513045776165,
+      "learning_rate": 1.947703941766411e-05,
+      "loss": 0.8092,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5666128770715946,
+      "learning_rate": 1.947598747529074e-05,
+      "loss": 0.7278,
+      "step": 1282
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6376187966620492,
+      "learning_rate": 1.947493450444712e-05,
+      "loss": 0.7695,
+      "step": 1283
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.406052287009246,
+      "learning_rate": 1.9473880505247532e-05,
+      "loss": 0.7641,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7269431831784723,
+      "learning_rate": 1.947282547780637e-05,
+      "loss": 0.8017,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5161138349707055,
+      "learning_rate": 1.947176942223814e-05,
+      "loss": 0.7157,
+      "step": 1286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5678540961023142,
+      "learning_rate": 1.947071233865746e-05,
+      "loss": 0.767,
+      "step": 1287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5527071981054203,
+      "learning_rate": 1.946965422717906e-05,
+      "loss": 0.8231,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5907878060262768,
+      "learning_rate": 1.9468595087917786e-05,
+      "loss": 0.6807,
+      "step": 1289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5158981766324053,
+      "learning_rate": 1.9467534920988586e-05,
+      "loss": 0.697,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.59870920201292,
+      "learning_rate": 1.9466473726506522e-05,
+      "loss": 0.7714,
+      "step": 1291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7794305184289632,
+      "learning_rate": 1.9465411504586778e-05,
+      "loss": 0.734,
+      "step": 1292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6392493833071595,
+      "learning_rate": 1.9464348255344642e-05,
+      "loss": 0.7791,
+      "step": 1293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6509845458827024,
+      "learning_rate": 1.9463283978895503e-05,
+      "loss": 0.7268,
+      "step": 1294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7374581302193655,
+      "learning_rate": 1.9462218675354877e-05,
+      "loss": 0.8593,
+      "step": 1295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5227355426200806,
+      "learning_rate": 1.9461152344838386e-05,
+      "loss": 0.766,
+      "step": 1296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.546737758110503,
+      "learning_rate": 1.9460084987461767e-05,
+      "loss": 0.7758,
+      "step": 1297
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5361827428965045,
+      "learning_rate": 1.945901660334086e-05,
+      "loss": 0.7971,
+      "step": 1298
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7399009926320979,
+      "learning_rate": 1.9457947192591623e-05,
+      "loss": 0.6787,
+      "step": 1299
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5621018072678603,
+      "learning_rate": 1.945687675533013e-05,
+      "loss": 0.7643,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5677030577309845,
+      "learning_rate": 1.9455805291672544e-05,
+      "loss": 0.7244,
+      "step": 1301
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8324847460117326,
+      "learning_rate": 1.9454732801735173e-05,
+      "loss": 0.8323,
+      "step": 1302
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8054740103348892,
+      "learning_rate": 1.945365928563441e-05,
+      "loss": 0.86,
+      "step": 1303
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5267902312222363,
+      "learning_rate": 1.9452584743486768e-05,
+      "loss": 0.7199,
+      "step": 1304
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6207235099791248,
+      "learning_rate": 1.9451509175408875e-05,
+      "loss": 0.8241,
+      "step": 1305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6311789356149866,
+      "learning_rate": 1.9450432581517466e-05,
+      "loss": 0.8855,
+      "step": 1306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6012394050190988,
+      "learning_rate": 1.944935496192939e-05,
+      "loss": 0.7032,
+      "step": 1307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6940041699409758,
+      "learning_rate": 1.9448276316761604e-05,
+      "loss": 0.795,
+      "step": 1308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6439835630134267,
+      "learning_rate": 1.9447196646131177e-05,
+      "loss": 0.8516,
+      "step": 1309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.563633479841595,
+      "learning_rate": 1.9446115950155293e-05,
+      "loss": 0.855,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6668575947195545,
+      "learning_rate": 1.9445034228951245e-05,
+      "loss": 0.7242,
+      "step": 1311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.573876033648897,
+      "learning_rate": 1.944395148263643e-05,
+      "loss": 0.7852,
+      "step": 1312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.77770874892001,
+      "learning_rate": 1.9442867711328372e-05,
+      "loss": 0.8421,
+      "step": 1313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6470437342507795,
+      "learning_rate": 1.9441782915144694e-05,
+      "loss": 0.762,
+      "step": 1314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5012094606088324,
+      "learning_rate": 1.9440697094203135e-05,
+      "loss": 0.7325,
+      "step": 1315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.624074833699309,
+      "learning_rate": 1.943961024862154e-05,
+      "loss": 0.7439,
+      "step": 1316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.853394164880661,
+      "learning_rate": 1.943852237851787e-05,
+      "loss": 0.7825,
+      "step": 1317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8797453479583488,
+      "learning_rate": 1.9437433484010203e-05,
+      "loss": 0.769,
+      "step": 1318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6375299522173172,
+      "learning_rate": 1.943634356521671e-05,
+      "loss": 0.7554,
+      "step": 1319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5538090573787875,
+      "learning_rate": 1.9435252622255694e-05,
+      "loss": 0.8113,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5303415868657724,
+      "learning_rate": 1.9434160655245557e-05,
+      "loss": 0.777,
+      "step": 1321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6464673627238309,
+      "learning_rate": 1.9433067664304818e-05,
+      "loss": 0.7287,
+      "step": 1322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.687666449674919,
+      "learning_rate": 1.9431973649552094e-05,
+      "loss": 0.7274,
+      "step": 1323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8097115392477354,
+      "learning_rate": 1.9430878611106133e-05,
+      "loss": 0.7362,
+      "step": 1324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5710783797225716,
+      "learning_rate": 1.942978254908578e-05,
+      "loss": 0.73,
+      "step": 1325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4434197275009946,
+      "learning_rate": 1.942868546361e-05,
+      "loss": 0.6966,
+      "step": 1326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7744289801449897,
+      "learning_rate": 1.9427587354797857e-05,
+      "loss": 0.8461,
+      "step": 1327
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5746501758483216,
+      "learning_rate": 1.9426488222768542e-05,
+      "loss": 0.7306,
+      "step": 1328
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6972017007527762,
+      "learning_rate": 1.942538806764134e-05,
+      "loss": 0.9094,
+      "step": 1329
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6418540267949544,
+      "learning_rate": 1.9424286889535663e-05,
+      "loss": 0.7503,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5414145441351583,
+      "learning_rate": 1.942318468857102e-05,
+      "loss": 0.809,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4873908203642654,
+      "learning_rate": 1.9422081464867043e-05,
+      "loss": 0.7576,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5279253204351688,
+      "learning_rate": 1.942097721854347e-05,
+      "loss": 0.7851,
+      "step": 1333
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.687191963317494,
+      "learning_rate": 1.9419871949720145e-05,
+      "loss": 0.8089,
+      "step": 1334
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4740624992934195,
+      "learning_rate": 1.9418765658517032e-05,
+      "loss": 0.8128,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.62078576616334,
+      "learning_rate": 1.9417658345054196e-05,
+      "loss": 0.7414,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4588705309609198,
+      "learning_rate": 1.9416550009451827e-05,
+      "loss": 0.6019,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5758310763195056,
+      "learning_rate": 1.941544065183021e-05,
+      "loss": 0.7815,
+      "step": 1338
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5708714704534004,
+      "learning_rate": 1.9414330272309753e-05,
+      "loss": 0.6871,
+      "step": 1339
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5364979929950109,
+      "learning_rate": 1.9413218871010964e-05,
+      "loss": 0.7195,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7385744275514265,
+      "learning_rate": 1.9412106448054476e-05,
+      "loss": 0.8279,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.731954304641849,
+      "learning_rate": 1.9410993003561027e-05,
+      "loss": 0.8354,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6744216970597119,
+      "learning_rate": 1.940987853765145e-05,
+      "loss": 0.7414,
+      "step": 1343
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6391277634441834,
+      "learning_rate": 1.9408763050446716e-05,
+      "loss": 0.7927,
+      "step": 1344
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7315724333267457,
+      "learning_rate": 1.9407646542067892e-05,
+      "loss": 0.8221,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6881225736292127,
+      "learning_rate": 1.9406529012636153e-05,
+      "loss": 0.7886,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.581344862065677,
+      "learning_rate": 1.940541046227279e-05,
+      "loss": 0.6217,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.589644667566598,
+      "learning_rate": 1.9404290891099205e-05,
+      "loss": 0.8205,
+      "step": 1348
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5314874491401882,
+      "learning_rate": 1.940317029923691e-05,
+      "loss": 0.7064,
+      "step": 1349
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6299195296123066,
+      "learning_rate": 1.9402048686807534e-05,
+      "loss": 0.6926,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6570600149638093,
+      "learning_rate": 1.94009260539328e-05,
+      "loss": 0.7882,
+      "step": 1351
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4966120899156123,
+      "learning_rate": 1.939980240073456e-05,
+      "loss": 0.8876,
+      "step": 1352
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5556698302348833,
+      "learning_rate": 1.939867772733476e-05,
+      "loss": 0.731,
+      "step": 1353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.733950508696902,
+      "learning_rate": 1.9397552033855474e-05,
+      "loss": 0.8104,
+      "step": 1354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6408561640363308,
+      "learning_rate": 1.939642532041888e-05,
+      "loss": 0.7891,
+      "step": 1355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6736879527988986,
+      "learning_rate": 1.9395297587147256e-05,
+      "loss": 0.8729,
+      "step": 1356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6896580180672505,
+      "learning_rate": 1.9394168834163007e-05,
+      "loss": 0.9141,
+      "step": 1357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6713377955139828,
+      "learning_rate": 1.939303906158864e-05,
+      "loss": 0.8107,
+      "step": 1358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6503043741757688,
+      "learning_rate": 1.9391908269546773e-05,
+      "loss": 0.8003,
+      "step": 1359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.47962070900276,
+      "learning_rate": 1.9390776458160137e-05,
+      "loss": 0.6367,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7702743050628662,
+      "learning_rate": 1.938964362755157e-05,
+      "loss": 0.8017,
+      "step": 1361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7824582581508035,
+      "learning_rate": 1.9388509777844027e-05,
+      "loss": 0.8004,
+      "step": 1362
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5668170478152565,
+      "learning_rate": 1.9387374909160567e-05,
+      "loss": 0.7007,
+      "step": 1363
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.560949917574267,
+      "learning_rate": 1.9386239021624362e-05,
+      "loss": 0.7871,
+      "step": 1364
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7083113146742042,
+      "learning_rate": 1.9385102115358695e-05,
+      "loss": 0.7337,
+      "step": 1365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6388928807175198,
+      "learning_rate": 1.938396419048696e-05,
+      "loss": 0.7306,
+      "step": 1366
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8150081650050511,
+      "learning_rate": 1.9382825247132662e-05,
+      "loss": 0.7636,
+      "step": 1367
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6574189204465755,
+      "learning_rate": 1.9381685285419417e-05,
+      "loss": 0.7559,
+      "step": 1368
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6478090214963608,
+      "learning_rate": 1.9380544305470942e-05,
+      "loss": 0.868,
+      "step": 1369
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.585684010108357,
+      "learning_rate": 1.937940230741108e-05,
+      "loss": 0.6967,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5524270780817286,
+      "learning_rate": 1.9378259291363774e-05,
+      "loss": 0.7274,
+      "step": 1371
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7849853674264955,
+      "learning_rate": 1.9377115257453085e-05,
+      "loss": 0.7452,
+      "step": 1372
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5268152637742454,
+      "learning_rate": 1.9375970205803175e-05,
+      "loss": 0.7099,
+      "step": 1373
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5911324780952263,
+      "learning_rate": 1.937482413653832e-05,
+      "loss": 0.8437,
+      "step": 1374
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4835029992058266,
+      "learning_rate": 1.9373677049782916e-05,
+      "loss": 0.5901,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5101360380804878,
+      "learning_rate": 1.937252894566145e-05,
+      "loss": 0.7386,
+      "step": 1376
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4448014915967933,
+      "learning_rate": 1.937137982429854e-05,
+      "loss": 0.7943,
+      "step": 1377
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5081393545005264,
+      "learning_rate": 1.9370229685818903e-05,
+      "loss": 0.663,
+      "step": 1378
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5379845123993552,
+      "learning_rate": 1.936907853034737e-05,
+      "loss": 0.7651,
+      "step": 1379
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7229609662472323,
+      "learning_rate": 1.9367926358008872e-05,
+      "loss": 0.624,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5857171073643888,
+      "learning_rate": 1.936677316892847e-05,
+      "loss": 0.7273,
+      "step": 1381
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8265782778745185,
+      "learning_rate": 1.936561896323132e-05,
+      "loss": 0.831,
+      "step": 1382
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5858906216474162,
+      "learning_rate": 1.9364463741042694e-05,
+      "loss": 0.8427,
+      "step": 1383
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6444591766781325,
+      "learning_rate": 1.936330750248797e-05,
+      "loss": 0.7696,
+      "step": 1384
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7707256531790037,
+      "learning_rate": 1.9362150247692646e-05,
+      "loss": 0.7645,
+      "step": 1385
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.510474176362667,
+      "learning_rate": 1.9360991976782317e-05,
+      "loss": 0.5463,
+      "step": 1386
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.616219735344741,
+      "learning_rate": 1.93598326898827e-05,
+      "loss": 0.6828,
+      "step": 1387
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5900808236080102,
+      "learning_rate": 1.935867238711962e-05,
+      "loss": 0.7644,
+      "step": 1388
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5447266911351711,
+      "learning_rate": 1.9357511068619e-05,
+      "loss": 0.7823,
+      "step": 1389
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.443662221459821,
+      "learning_rate": 1.9356348734506888e-05,
+      "loss": 0.6705,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6858155970178428,
+      "learning_rate": 1.935518538490944e-05,
+      "loss": 0.7775,
+      "step": 1391
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6260463865702544,
+      "learning_rate": 1.9354021019952917e-05,
+      "loss": 0.6997,
+      "step": 1392
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6151462386185187,
+      "learning_rate": 1.9352855639763693e-05,
+      "loss": 0.7388,
+      "step": 1393
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6226745231864121,
+      "learning_rate": 1.935168924446825e-05,
+      "loss": 0.7637,
+      "step": 1394
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.736869739132547,
+      "learning_rate": 1.935052183419319e-05,
+      "loss": 0.8231,
+      "step": 1395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5521494478169477,
+      "learning_rate": 1.9349353409065203e-05,
+      "loss": 0.8538,
+      "step": 1396
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4936808683766414,
+      "learning_rate": 1.9348183969211113e-05,
+      "loss": 0.7137,
+      "step": 1397
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6405737267346916,
+      "learning_rate": 1.9347013514757845e-05,
+      "loss": 0.7905,
+      "step": 1398
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.621940251119487,
+      "learning_rate": 1.934584204583243e-05,
+      "loss": 0.828,
+      "step": 1399
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6037284742563462,
+      "learning_rate": 1.934466956256201e-05,
+      "loss": 0.6796,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.567145940022025,
+      "learning_rate": 1.9343496065073846e-05,
+      "loss": 0.7317,
+      "step": 1401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9771936077668393,
+      "learning_rate": 1.93423215534953e-05,
+      "loss": 0.7907,
+      "step": 1402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6312584772073946,
+      "learning_rate": 1.934114602795385e-05,
+      "loss": 0.6874,
+      "step": 1403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4670109880400073,
+      "learning_rate": 1.9339969488577074e-05,
+      "loss": 0.7118,
+      "step": 1404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5236031544065936,
+      "learning_rate": 1.933879193549267e-05,
+      "loss": 0.7507,
+      "step": 1405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6547095135909067,
+      "learning_rate": 1.9337613368828443e-05,
+      "loss": 0.7915,
+      "step": 1406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6994751897935607,
+      "learning_rate": 1.9336433788712313e-05,
+      "loss": 0.7358,
+      "step": 1407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6317326459822472,
+      "learning_rate": 1.9335253195272298e-05,
+      "loss": 0.7229,
+      "step": 1408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4262997590885145,
+      "learning_rate": 1.9334071588636537e-05,
+      "loss": 0.6226,
+      "step": 1409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.650951039532934,
+      "learning_rate": 1.9332888968933273e-05,
+      "loss": 0.7597,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.431463986073521,
+      "learning_rate": 1.933170533629086e-05,
+      "loss": 0.7399,
+      "step": 1411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7834899114222995,
+      "learning_rate": 1.933052069083777e-05,
+      "loss": 0.7878,
+      "step": 1412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6821249271791365,
+      "learning_rate": 1.9329335032702564e-05,
+      "loss": 0.7578,
+      "step": 1413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6168283852908498,
+      "learning_rate": 1.932814836201394e-05,
+      "loss": 0.8011,
+      "step": 1414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5598551085056578,
+      "learning_rate": 1.932696067890069e-05,
+      "loss": 0.7733,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4618766523100533,
+      "learning_rate": 1.9325771983491708e-05,
+      "loss": 0.6826,
+      "step": 1416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3812200411588107,
+      "learning_rate": 1.932458227591602e-05,
+      "loss": 0.7218,
+      "step": 1417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5380172482866934,
+      "learning_rate": 1.9323391556302743e-05,
+      "loss": 0.8418,
+      "step": 1418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6412778862474402,
+      "learning_rate": 1.9322199824781117e-05,
+      "loss": 0.826,
+      "step": 1419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5275748605652015,
+      "learning_rate": 1.9321007081480486e-05,
+      "loss": 0.7854,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4810290185183532,
+      "learning_rate": 1.9319813326530296e-05,
+      "loss": 0.7001,
+      "step": 1421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5757069867975266,
+      "learning_rate": 1.9318618560060117e-05,
+      "loss": 0.7207,
+      "step": 1422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4581775676110735,
+      "learning_rate": 1.931742278219962e-05,
+      "loss": 0.7473,
+      "step": 1423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5452482775653402,
+      "learning_rate": 1.9316225993078586e-05,
+      "loss": 0.7197,
+      "step": 1424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5196283043899141,
+      "learning_rate": 1.9315028192826912e-05,
+      "loss": 0.7792,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.490523064687223,
+      "learning_rate": 1.9313829381574594e-05,
+      "loss": 0.714,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5986736451592876,
+      "learning_rate": 1.9312629559451755e-05,
+      "loss": 0.8192,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6455596522866387,
+      "learning_rate": 1.931142872658861e-05,
+      "loss": 0.8674,
+      "step": 1428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.690847570583257,
+      "learning_rate": 1.9310226883115488e-05,
+      "loss": 0.7916,
+      "step": 1429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7743202508358853,
+      "learning_rate": 1.9309024029162833e-05,
+      "loss": 0.5859,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5466754395817315,
+      "learning_rate": 1.93078201648612e-05,
+      "loss": 0.7705,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5247144984670906,
+      "learning_rate": 1.9306615290341244e-05,
+      "loss": 0.7685,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6077323413863682,
+      "learning_rate": 1.9305409405733736e-05,
+      "loss": 0.7231,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6702697638591097,
+      "learning_rate": 1.930420251116956e-05,
+      "loss": 0.7152,
+      "step": 1434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4667964469637342,
+      "learning_rate": 1.9302994606779704e-05,
+      "loss": 0.6875,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.525503816062669,
+      "learning_rate": 1.930178569269526e-05,
+      "loss": 0.684,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8024932861274463,
+      "learning_rate": 1.9300575769047448e-05,
+      "loss": 0.7442,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.562052851367724,
+      "learning_rate": 1.929936483596758e-05,
+      "loss": 0.7246,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6614152158765623,
+      "learning_rate": 1.9298152893587084e-05,
+      "loss": 0.7837,
+      "step": 1439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7276983476208452,
+      "learning_rate": 1.92969399420375e-05,
+      "loss": 0.732,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5647712919266812,
+      "learning_rate": 1.9295725981450473e-05,
+      "loss": 0.7849,
+      "step": 1441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6277974927232288,
+      "learning_rate": 1.9294511011957757e-05,
+      "loss": 0.7406,
+      "step": 1442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6582286211160586,
+      "learning_rate": 1.9293295033691223e-05,
+      "loss": 0.7728,
+      "step": 1443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.581468131712894,
+      "learning_rate": 1.9292078046782844e-05,
+      "loss": 0.7779,
+      "step": 1444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7121934073770229,
+      "learning_rate": 1.9290860051364704e-05,
+      "loss": 0.8609,
+      "step": 1445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.549302609178488,
+      "learning_rate": 1.9289641047569e-05,
+      "loss": 0.6958,
+      "step": 1446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6043051626537128,
+      "learning_rate": 1.928842103552803e-05,
+      "loss": 0.7551,
+      "step": 1447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6362496985196373,
+      "learning_rate": 1.928720001537421e-05,
+      "loss": 0.8159,
+      "step": 1448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8244786394221024,
+      "learning_rate": 1.928597798724007e-05,
+      "loss": 0.8213,
+      "step": 1449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.670704556030598,
+      "learning_rate": 1.9284754951258233e-05,
+      "loss": 0.8299,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.511410122208224,
+      "learning_rate": 1.9283530907561445e-05,
+      "loss": 0.7247,
+      "step": 1451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6005491815394173,
+      "learning_rate": 1.9282305856282554e-05,
+      "loss": 0.7206,
+      "step": 1452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.623120296961341,
+      "learning_rate": 1.928107979755452e-05,
+      "loss": 0.669,
+      "step": 1453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7108125044473872,
+      "learning_rate": 1.9279852731510417e-05,
+      "loss": 0.752,
+      "step": 1454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5740653272856195,
+      "learning_rate": 1.927862465828342e-05,
+      "loss": 0.7685,
+      "step": 1455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5311910116271827,
+      "learning_rate": 1.9277395578006815e-05,
+      "loss": 0.7808,
+      "step": 1456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7568559916243593,
+      "learning_rate": 1.9276165490814e-05,
+      "loss": 0.7578,
+      "step": 1457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6287705974238318,
+      "learning_rate": 1.927493439683849e-05,
+      "loss": 0.7119,
+      "step": 1458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7208541800835253,
+      "learning_rate": 1.9273702296213893e-05,
+      "loss": 0.8291,
+      "step": 1459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5970064188840418,
+      "learning_rate": 1.9272469189073936e-05,
+      "loss": 0.7491,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.591605260808786,
+      "learning_rate": 1.9271235075552454e-05,
+      "loss": 0.6758,
+      "step": 1461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5906047294247754,
+      "learning_rate": 1.926999995578339e-05,
+      "loss": 0.7827,
+      "step": 1462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8630511855892309,
+      "learning_rate": 1.9268763829900798e-05,
+      "loss": 0.7945,
+      "step": 1463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.793429611608883,
+      "learning_rate": 1.9267526698038838e-05,
+      "loss": 0.8592,
+      "step": 1464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.803266576136559,
+      "learning_rate": 1.9266288560331782e-05,
+      "loss": 0.7465,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5244435320341803,
+      "learning_rate": 1.9265049416914015e-05,
+      "loss": 0.7006,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7823386142447417,
+      "learning_rate": 1.9263809267920023e-05,
+      "loss": 0.8138,
+      "step": 1467
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.546354254225523,
+      "learning_rate": 1.92625681134844e-05,
+      "loss": 0.6401,
+      "step": 1468
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5043994110416876,
+      "learning_rate": 1.9261325953741863e-05,
+      "loss": 0.8051,
+      "step": 1469
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6999701502405948,
+      "learning_rate": 1.9260082788827225e-05,
+      "loss": 0.8232,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6116709705664414,
+      "learning_rate": 1.9258838618875407e-05,
+      "loss": 0.7465,
+      "step": 1471
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5700867235311593,
+      "learning_rate": 1.9257593444021455e-05,
+      "loss": 0.6726,
+      "step": 1472
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6404232682439661,
+      "learning_rate": 1.92563472644005e-05,
+      "loss": 0.8057,
+      "step": 1473
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.609198643059564,
+      "learning_rate": 1.9255100080147807e-05,
+      "loss": 0.8679,
+      "step": 1474
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.61013255267516,
+      "learning_rate": 1.9253851891398735e-05,
+      "loss": 0.732,
+      "step": 1475
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5979956087776157,
+      "learning_rate": 1.9252602698288752e-05,
+      "loss": 0.7424,
+      "step": 1476
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5554224156137497,
+      "learning_rate": 1.925135250095344e-05,
+      "loss": 0.7008,
+      "step": 1477
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6167967536593908,
+      "learning_rate": 1.9250101299528495e-05,
+      "loss": 0.8089,
+      "step": 1478
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6375834796003947,
+      "learning_rate": 1.9248849094149703e-05,
+      "loss": 0.7496,
+      "step": 1479
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5682760543403722,
+      "learning_rate": 1.9247595884952977e-05,
+      "loss": 0.8036,
+      "step": 1480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6028509146470817,
+      "learning_rate": 1.924634167207434e-05,
+      "loss": 0.7174,
+      "step": 1481
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6240045575203728,
+      "learning_rate": 1.9245086455649903e-05,
+      "loss": 0.7451,
+      "step": 1482
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.619210407596221,
+      "learning_rate": 1.9243830235815913e-05,
+      "loss": 0.7987,
+      "step": 1483
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4698223414285452,
+      "learning_rate": 1.924257301270871e-05,
+      "loss": 0.7055,
+      "step": 1484
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5073455061374148,
+      "learning_rate": 1.924131478646474e-05,
+      "loss": 0.8317,
+      "step": 1485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6407134053525472,
+      "learning_rate": 1.9240055557220573e-05,
+      "loss": 0.7548,
+      "step": 1486
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8031573777028704,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 0.7956,
+      "step": 1487
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.582598949302052,
+      "learning_rate": 1.923753409027841e-05,
+      "loss": 0.7401,
+      "step": 1488
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4703010435992137,
+      "learning_rate": 1.9236271852854088e-05,
+      "loss": 0.8035,
+      "step": 1489
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4954150702770912,
+      "learning_rate": 1.9235008612976897e-05,
+      "loss": 0.7635,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9250287248373963,
+      "learning_rate": 1.9233744370783935e-05,
+      "loss": 0.7855,
+      "step": 1491
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6460987951004957,
+      "learning_rate": 1.9232479126412425e-05,
+      "loss": 0.8874,
+      "step": 1492
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.563877453344683,
+      "learning_rate": 1.9231212879999686e-05,
+      "loss": 0.8022,
+      "step": 1493
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7544917369182025,
+      "learning_rate": 1.922994563168315e-05,
+      "loss": 0.7461,
+      "step": 1494
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6902299666720255,
+      "learning_rate": 1.9228677381600352e-05,
+      "loss": 0.7271,
+      "step": 1495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.672354406035893,
+      "learning_rate": 1.9227408129888945e-05,
+      "loss": 0.7699,
+      "step": 1496
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.396315478816546,
+      "learning_rate": 1.9226137876686686e-05,
+      "loss": 0.7364,
+      "step": 1497
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7579916394018387,
+      "learning_rate": 1.9224866622131442e-05,
+      "loss": 0.7713,
+      "step": 1498
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5314102029803376,
+      "learning_rate": 1.9223594366361188e-05,
+      "loss": 0.8544,
+      "step": 1499
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5980160334993416,
+      "learning_rate": 1.9222321109514006e-05,
+      "loss": 0.8614,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5790504981956814,
+      "learning_rate": 1.922104685172809e-05,
+      "loss": 0.6682,
+      "step": 1501
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.506221154074456,
+      "learning_rate": 1.9219771593141736e-05,
+      "loss": 0.827,
+      "step": 1502
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6994180118450408,
+      "learning_rate": 1.921849533389336e-05,
+      "loss": 0.7882,
+      "step": 1503
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5998130523759928,
+      "learning_rate": 1.9217218074121474e-05,
+      "loss": 0.6596,
+      "step": 1504
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.524449341063304,
+      "learning_rate": 1.9215939813964707e-05,
+      "loss": 0.7048,
+      "step": 1505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5581478522339887,
+      "learning_rate": 1.9214660553561796e-05,
+      "loss": 0.7535,
+      "step": 1506
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.616475689899037,
+      "learning_rate": 1.921338029305158e-05,
+      "loss": 0.8237,
+      "step": 1507
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.540256961106526,
+      "learning_rate": 1.9212099032573017e-05,
+      "loss": 0.7855,
+      "step": 1508
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6530635115493926,
+      "learning_rate": 1.9210816772265166e-05,
+      "loss": 0.7577,
+      "step": 1509
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6069626479514616,
+      "learning_rate": 1.9209533512267193e-05,
+      "loss": 0.793,
+      "step": 1510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.726638072569584,
+      "learning_rate": 1.920824925271838e-05,
+      "loss": 0.8787,
+      "step": 1511
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.622170668313893,
+      "learning_rate": 1.920696399375811e-05,
+      "loss": 0.6719,
+      "step": 1512
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5849164974116032,
+      "learning_rate": 1.9205677735525883e-05,
+      "loss": 0.7252,
+      "step": 1513
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6194613219944933,
+      "learning_rate": 1.9204390478161294e-05,
+      "loss": 0.8939,
+      "step": 1514
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5088878684059277,
+      "learning_rate": 1.9203102221804062e-05,
+      "loss": 0.7617,
+      "step": 1515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7035212373349042,
+      "learning_rate": 1.9201812966594e-05,
+      "loss": 0.7593,
+      "step": 1516
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6703938590394398,
+      "learning_rate": 1.920052271267105e-05,
+      "loss": 0.6431,
+      "step": 1517
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5769045147068257,
+      "learning_rate": 1.9199231460175232e-05,
+      "loss": 0.6884,
+      "step": 1518
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5809397366575757,
+      "learning_rate": 1.9197939209246697e-05,
+      "loss": 0.7772,
+      "step": 1519
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6391801398503676,
+      "learning_rate": 1.9196645960025707e-05,
+      "loss": 0.7642,
+      "step": 1520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8299512941006664,
+      "learning_rate": 1.9195351712652615e-05,
+      "loss": 0.7433,
+      "step": 1521
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5095893517785566,
+      "learning_rate": 1.9194056467267893e-05,
+      "loss": 0.7138,
+      "step": 1522
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6093122134592646,
+      "learning_rate": 1.919276022401212e-05,
+      "loss": 0.8123,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6161664662430497,
+      "learning_rate": 1.9191462983025984e-05,
+      "loss": 0.7022,
+      "step": 1524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.761508380087029,
+      "learning_rate": 1.919016474445028e-05,
+      "loss": 0.7325,
+      "step": 1525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6202857694097255,
+      "learning_rate": 1.9188865508425912e-05,
+      "loss": 0.7897,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.566351498765691,
+      "learning_rate": 1.918756527509389e-05,
+      "loss": 0.7244,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5388170986926715,
+      "learning_rate": 1.9186264044595334e-05,
+      "loss": 0.7408,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0411475883686667,
+      "learning_rate": 1.9184961817071474e-05,
+      "loss": 0.8856,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.537923172996495,
+      "learning_rate": 1.9183658592663646e-05,
+      "loss": 0.7629,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5649059968966519,
+      "learning_rate": 1.918235437151329e-05,
+      "loss": 0.8007,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4841845007351928,
+      "learning_rate": 1.9181049153761968e-05,
+      "loss": 0.7615,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5810852008291814,
+      "learning_rate": 1.9179742939551334e-05,
+      "loss": 0.7283,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6433971912621534,
+      "learning_rate": 1.917843572902316e-05,
+      "loss": 0.7719,
+      "step": 1534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6686771191429626,
+      "learning_rate": 1.917712752231932e-05,
+      "loss": 0.7082,
+      "step": 1535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5992105637180958,
+      "learning_rate": 1.9175818319581804e-05,
+      "loss": 0.8333,
+      "step": 1536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5199822157361877,
+      "learning_rate": 1.9174508120952703e-05,
+      "loss": 0.6218,
+      "step": 1537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.696677016341126,
+      "learning_rate": 1.9173196926574217e-05,
+      "loss": 0.774,
+      "step": 1538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5505851914086217,
+      "learning_rate": 1.9171884736588658e-05,
+      "loss": 0.8452,
+      "step": 1539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6196310318726466,
+      "learning_rate": 1.9170571551138443e-05,
+      "loss": 0.7345,
+      "step": 1540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.711535964064723,
+      "learning_rate": 1.91692573703661e-05,
+      "loss": 0.8005,
+      "step": 1541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.602974087584989,
+      "learning_rate": 1.916794219441426e-05,
+      "loss": 0.7703,
+      "step": 1542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.803421607568332,
+      "learning_rate": 1.9166626023425663e-05,
+      "loss": 0.7639,
+      "step": 1543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6301320994866801,
+      "learning_rate": 1.9165308857543162e-05,
+      "loss": 0.7741,
+      "step": 1544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.923659935329634,
+      "learning_rate": 1.9163990696909714e-05,
+      "loss": 0.83,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.833057279021541,
+      "learning_rate": 1.9162671541668384e-05,
+      "loss": 0.8549,
+      "step": 1546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5499426415563595,
+      "learning_rate": 1.9161351391962347e-05,
+      "loss": 0.7802,
+      "step": 1547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.61928782607135,
+      "learning_rate": 1.9160030247934887e-05,
+      "loss": 0.8191,
+      "step": 1548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.454897568051772,
+      "learning_rate": 1.9158708109729387e-05,
+      "loss": 0.6563,
+      "step": 1549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6785044647038443,
+      "learning_rate": 1.9157384977489347e-05,
+      "loss": 0.7766,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7302652329541985,
+      "learning_rate": 1.9156060851358377e-05,
+      "loss": 0.709,
+      "step": 1551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4544831620456349,
+      "learning_rate": 1.9154735731480185e-05,
+      "loss": 0.6414,
+      "step": 1552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6795340410621473,
+      "learning_rate": 1.915340961799859e-05,
+      "loss": 0.8081,
+      "step": 1553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4616387300706588,
+      "learning_rate": 1.9152082511057525e-05,
+      "loss": 0.8863,
+      "step": 1554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.522043264691204,
+      "learning_rate": 1.915075441080103e-05,
+      "loss": 0.7951,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.567166649488013,
+      "learning_rate": 1.9149425317373242e-05,
+      "loss": 0.8348,
+      "step": 1556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.651202931472597,
+      "learning_rate": 1.914809523091842e-05,
+      "loss": 0.7499,
+      "step": 1557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6437075117820903,
+      "learning_rate": 1.9146764151580916e-05,
+      "loss": 0.7034,
+      "step": 1558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.955079683704659,
+      "learning_rate": 1.9145432079505205e-05,
+      "loss": 0.802,
+      "step": 1559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5064482039439775,
+      "learning_rate": 1.9144099014835863e-05,
+      "loss": 0.7824,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5106371142000838,
+      "learning_rate": 1.914276495771757e-05,
+      "loss": 0.6852,
+      "step": 1561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4133764252186334,
+      "learning_rate": 1.9141429908295115e-05,
+      "loss": 0.7393,
+      "step": 1562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5175862671123022,
+      "learning_rate": 1.91400938667134e-05,
+      "loss": 0.652,
+      "step": 1563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8668591354206043,
+      "learning_rate": 1.9138756833117433e-05,
+      "loss": 0.8072,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7148653540172474,
+      "learning_rate": 1.9137418807652322e-05,
+      "loss": 0.7976,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3971811107739365,
+      "learning_rate": 1.9136079790463297e-05,
+      "loss": 0.6894,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.558703271920691,
+      "learning_rate": 1.913473978169568e-05,
+      "loss": 0.7824,
+      "step": 1567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6105428293661352,
+      "learning_rate": 1.9133398781494915e-05,
+      "loss": 0.8303,
+      "step": 1568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.443076696573244,
+      "learning_rate": 1.9132056790006538e-05,
+      "loss": 0.6699,
+      "step": 1569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5389324497104773,
+      "learning_rate": 1.9130713807376208e-05,
+      "loss": 0.7501,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.418175886216392,
+      "learning_rate": 1.9129369833749684e-05,
+      "loss": 0.7416,
+      "step": 1571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5503745273715697,
+      "learning_rate": 1.9128024869272827e-05,
+      "loss": 0.8613,
+      "step": 1572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5373599502009552,
+      "learning_rate": 1.912667891409162e-05,
+      "loss": 0.6855,
+      "step": 1573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.778854879435199,
+      "learning_rate": 1.9125331968352144e-05,
+      "loss": 0.838,
+      "step": 1574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6617599172340474,
+      "learning_rate": 1.9123984032200586e-05,
+      "loss": 0.8029,
+      "step": 1575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5626142896457125,
+      "learning_rate": 1.9122635105783246e-05,
+      "loss": 0.7714,
+      "step": 1576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5953250529453096,
+      "learning_rate": 1.9121285189246522e-05,
+      "loss": 0.7895,
+      "step": 1577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8414569875633697,
+      "learning_rate": 1.9119934282736937e-05,
+      "loss": 0.9218,
+      "step": 1578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4706065042563319,
+      "learning_rate": 1.91185823864011e-05,
+      "loss": 0.6972,
+      "step": 1579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5299081578716949,
+      "learning_rate": 1.9117229500385747e-05,
+      "loss": 0.7931,
+      "step": 1580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3789090369504793,
+      "learning_rate": 1.9115875624837712e-05,
+      "loss": 0.7851,
+      "step": 1581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4746021435408434,
+      "learning_rate": 1.911452075990393e-05,
+      "loss": 0.8104,
+      "step": 1582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6450850823417302,
+      "learning_rate": 1.9113164905731456e-05,
+      "loss": 0.7881,
+      "step": 1583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5010962716226026,
+      "learning_rate": 1.9111808062467444e-05,
+      "loss": 0.7363,
+      "step": 1584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6404627346035592,
+      "learning_rate": 1.9110450230259164e-05,
+      "loss": 0.7816,
+      "step": 1585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6060822901587557,
+      "learning_rate": 1.910909140925398e-05,
+      "loss": 0.8176,
+      "step": 1586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.51717087233064,
+      "learning_rate": 1.910773159959937e-05,
+      "loss": 0.7383,
+      "step": 1587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.61362295729002,
+      "learning_rate": 1.9106370801442932e-05,
+      "loss": 0.7415,
+      "step": 1588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4683775876196694,
+      "learning_rate": 1.9105009014932346e-05,
+      "loss": 0.6588,
+      "step": 1589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.770666499944069,
+      "learning_rate": 1.9103646240215417e-05,
+      "loss": 0.7953,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.738033048483459,
+      "learning_rate": 1.9102282477440054e-05,
+      "loss": 0.7054,
+      "step": 1591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5252116881641011,
+      "learning_rate": 1.9100917726754274e-05,
+      "loss": 0.8349,
+      "step": 1592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.628771819360274,
+      "learning_rate": 1.9099551988306196e-05,
+      "loss": 0.7939,
+      "step": 1593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5997324728092637,
+      "learning_rate": 1.9098185262244052e-05,
+      "loss": 0.7922,
+      "step": 1594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.577128392588311,
+      "learning_rate": 1.9096817548716176e-05,
+      "loss": 0.7836,
+      "step": 1595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7620820504603598,
+      "learning_rate": 1.9095448847871017e-05,
+      "loss": 0.6764,
+      "step": 1596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4430637543169647,
+      "learning_rate": 1.909407915985712e-05,
+      "loss": 0.7046,
+      "step": 1597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5525390305375857,
+      "learning_rate": 1.9092708484823146e-05,
+      "loss": 0.778,
+      "step": 1598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6033162887065793,
+      "learning_rate": 1.9091336822917858e-05,
+      "loss": 0.734,
+      "step": 1599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.373049548183506,
+      "learning_rate": 1.908996417429013e-05,
+      "loss": 0.7226,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.515664936780327,
+      "learning_rate": 1.9088590539088944e-05,
+      "loss": 0.8194,
+      "step": 1601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.676295501712208,
+      "learning_rate": 1.9087215917463388e-05,
+      "loss": 0.7845,
+      "step": 1602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7349913040217095,
+      "learning_rate": 1.9085840309562647e-05,
+      "loss": 0.7594,
+      "step": 1603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4118887464545329,
+      "learning_rate": 1.9084463715536028e-05,
+      "loss": 0.7238,
+      "step": 1604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5514927958243494,
+      "learning_rate": 1.908308613553294e-05,
+      "loss": 0.8336,
+      "step": 1605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7937303832748122,
+      "learning_rate": 1.9081707569702895e-05,
+      "loss": 0.8283,
+      "step": 1606
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6048560592421042,
+      "learning_rate": 1.9080328018195512e-05,
+      "loss": 0.9193,
+      "step": 1607
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5251342315015357,
+      "learning_rate": 1.9078947481160525e-05,
+      "loss": 0.7064,
+      "step": 1608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5915408336271724,
+      "learning_rate": 1.907756595874777e-05,
+      "loss": 0.8293,
+      "step": 1609
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5361474379210114,
+      "learning_rate": 1.9076183451107185e-05,
+      "loss": 0.7543,
+      "step": 1610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6348641221810145,
+      "learning_rate": 1.9074799958388824e-05,
+      "loss": 0.8217,
+      "step": 1611
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6079219437209098,
+      "learning_rate": 1.907341548074284e-05,
+      "loss": 0.8163,
+      "step": 1612
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5291210741782348,
+      "learning_rate": 1.9072030018319498e-05,
+      "loss": 0.8774,
+      "step": 1613
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6444665406797316,
+      "learning_rate": 1.9070643571269168e-05,
+      "loss": 0.7758,
+      "step": 1614
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4431843276318772,
+      "learning_rate": 1.906925613974233e-05,
+      "loss": 0.6653,
+      "step": 1615
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.567521467193932,
+      "learning_rate": 1.9067867723889563e-05,
+      "loss": 0.8517,
+      "step": 1616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5714406321988241,
+      "learning_rate": 1.9066478323861558e-05,
+      "loss": 0.7835,
+      "step": 1617
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5959228661639966,
+      "learning_rate": 1.906508793980912e-05,
+      "loss": 0.7141,
+      "step": 1618
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4847876302277403,
+      "learning_rate": 1.9063696571883145e-05,
+      "loss": 0.7785,
+      "step": 1619
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5503075253886878,
+      "learning_rate": 1.9062304220234648e-05,
+      "loss": 0.7935,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6556741264913504,
+      "learning_rate": 1.9060910885014745e-05,
+      "loss": 0.6974,
+      "step": 1621
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4290189452587512,
+      "learning_rate": 1.9059516566374662e-05,
+      "loss": 0.6936,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5478177845334817,
+      "learning_rate": 1.9058121264465733e-05,
+      "loss": 0.7785,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5658364846871906,
+      "learning_rate": 1.9056724979439395e-05,
+      "loss": 0.8645,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5477169944101723,
+      "learning_rate": 1.9055327711447188e-05,
+      "loss": 0.7858,
+      "step": 1625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4657396785160544,
+      "learning_rate": 1.905392946064077e-05,
+      "loss": 0.7117,
+      "step": 1626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5708545119734512,
+      "learning_rate": 1.90525302271719e-05,
+      "loss": 0.7021,
+      "step": 1627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6476076431039985,
+      "learning_rate": 1.9051130011192432e-05,
+      "loss": 0.7394,
+      "step": 1628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4400462499040607,
+      "learning_rate": 1.904972881285435e-05,
+      "loss": 0.8131,
+      "step": 1629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6174406456486523,
+      "learning_rate": 1.9048326632309725e-05,
+      "loss": 0.7308,
+      "step": 1630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5193393003943425,
+      "learning_rate": 1.9046923469710745e-05,
+      "loss": 0.7133,
+      "step": 1631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7774121870566917,
+      "learning_rate": 1.90455193252097e-05,
+      "loss": 0.8559,
+      "step": 1632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.526356917860105,
+      "learning_rate": 1.904411419895899e-05,
+      "loss": 0.6728,
+      "step": 1633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.869316998443361,
+      "learning_rate": 1.9042708091111118e-05,
+      "loss": 0.8109,
+      "step": 1634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6306679434589662,
+      "learning_rate": 1.9041301001818694e-05,
+      "loss": 0.7945,
+      "step": 1635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.757448536351054,
+      "learning_rate": 1.9039892931234434e-05,
+      "loss": 0.7686,
+      "step": 1636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.544620177922369,
+      "learning_rate": 1.9038483879511166e-05,
+      "loss": 0.6817,
+      "step": 1637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7468142129124404,
+      "learning_rate": 1.903707384680182e-05,
+      "loss": 0.7455,
+      "step": 1638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6848255747831735,
+      "learning_rate": 1.9035662833259433e-05,
+      "loss": 0.8196,
+      "step": 1639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5329377292764257,
+      "learning_rate": 1.9034250839037144e-05,
+      "loss": 0.7906,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5415068985564901,
+      "learning_rate": 1.903283786428821e-05,
+      "loss": 0.6799,
+      "step": 1641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4666601543044273,
+      "learning_rate": 1.903142390916598e-05,
+      "loss": 0.7268,
+      "step": 1642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.575405053556528,
+      "learning_rate": 1.903000897382392e-05,
+      "loss": 0.7831,
+      "step": 1643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6124679892613942,
+      "learning_rate": 1.9028593058415604e-05,
+      "loss": 0.7744,
+      "step": 1644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4810041986168434,
+      "learning_rate": 1.90271761630947e-05,
+      "loss": 0.7654,
+      "step": 1645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5817930470576074,
+      "learning_rate": 1.9025758288014992e-05,
+      "loss": 0.7471,
+      "step": 1646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7806556950470092,
+      "learning_rate": 1.9024339433330374e-05,
+      "loss": 0.7575,
+      "step": 1647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7388350084069517,
+      "learning_rate": 1.902291959919483e-05,
+      "loss": 0.7252,
+      "step": 1648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6049185515811917,
+      "learning_rate": 1.9021498785762465e-05,
+      "loss": 0.71,
+      "step": 1649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5496687398901938,
+      "learning_rate": 1.902007699318749e-05,
+      "loss": 0.8395,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5859664455327087,
+      "learning_rate": 1.9018654221624215e-05,
+      "loss": 0.8465,
+      "step": 1651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6860476567662297,
+      "learning_rate": 1.9017230471227065e-05,
+      "loss": 0.8093,
+      "step": 1652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5255507157557995,
+      "learning_rate": 1.9015805742150555e-05,
+      "loss": 0.7641,
+      "step": 1653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5576595287920239,
+      "learning_rate": 1.9014380034549326e-05,
+      "loss": 0.7399,
+      "step": 1654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5716277986521965,
+      "learning_rate": 1.901295334857811e-05,
+      "loss": 0.7475,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5940348916804257,
+      "learning_rate": 1.901152568439176e-05,
+      "loss": 0.8035,
+      "step": 1656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.652806276890107,
+      "learning_rate": 1.9010097042145217e-05,
+      "loss": 0.7669,
+      "step": 1657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5652094056502321,
+      "learning_rate": 1.9008667421993542e-05,
+      "loss": 0.7551,
+      "step": 1658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4934518396928138,
+      "learning_rate": 1.9007236824091902e-05,
+      "loss": 0.8183,
+      "step": 1659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4623729202198175,
+      "learning_rate": 1.9005805248595558e-05,
+      "loss": 0.7898,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.572320496115777,
+      "learning_rate": 1.9004372695659893e-05,
+      "loss": 0.7187,
+      "step": 1661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5236632692717045,
+      "learning_rate": 1.9002939165440382e-05,
+      "loss": 0.7906,
+      "step": 1662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6567306846632446,
+      "learning_rate": 1.9001504658092614e-05,
+      "loss": 0.7019,
+      "step": 1663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7033509945844678,
+      "learning_rate": 1.900006917377229e-05,
+      "loss": 0.8409,
+      "step": 1664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5112174633108457,
+      "learning_rate": 1.8998632712635197e-05,
+      "loss": 0.6675,
+      "step": 1665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6077579566638796,
+      "learning_rate": 1.8997195274837246e-05,
+      "loss": 0.7659,
+      "step": 1666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6200552954296608,
+      "learning_rate": 1.8995756860534447e-05,
+      "loss": 0.7704,
+      "step": 1667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7219390404142412,
+      "learning_rate": 1.899431746988292e-05,
+      "loss": 0.6929,
+      "step": 1668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5052480569089988,
+      "learning_rate": 1.8992877103038888e-05,
+      "loss": 0.7769,
+      "step": 1669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7106222019790918,
+      "learning_rate": 1.8991435760158678e-05,
+      "loss": 0.7939,
+      "step": 1670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6093050462228884,
+      "learning_rate": 1.8989993441398725e-05,
+      "loss": 0.7808,
+      "step": 1671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6456229403185043,
+      "learning_rate": 1.8988550146915577e-05,
+      "loss": 0.7685,
+      "step": 1672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6032421563969286,
+      "learning_rate": 1.8987105876865875e-05,
+      "loss": 0.769,
+      "step": 1673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.583807363251475,
+      "learning_rate": 1.898566063140637e-05,
+      "loss": 0.7032,
+      "step": 1674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.493731079727449,
+      "learning_rate": 1.898421441069392e-05,
+      "loss": 0.7107,
+      "step": 1675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5749307014664113,
+      "learning_rate": 1.89827672148855e-05,
+      "loss": 0.8638,
+      "step": 1676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6771481017193612,
+      "learning_rate": 1.898131904413817e-05,
+      "loss": 0.7711,
+      "step": 1677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4599515555865261,
+      "learning_rate": 1.897986989860911e-05,
+      "loss": 0.7718,
+      "step": 1678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.748238922772858,
+      "learning_rate": 1.8978419778455604e-05,
+      "loss": 0.7038,
+      "step": 1679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7105558563254843,
+      "learning_rate": 1.8976968683835035e-05,
+      "loss": 0.7343,
+      "step": 1680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.53133673055251,
+      "learning_rate": 1.89755166149049e-05,
+      "loss": 0.7529,
+      "step": 1681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5783781823685694,
+      "learning_rate": 1.8974063571822802e-05,
+      "loss": 0.8669,
+      "step": 1682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.657796324266244,
+      "learning_rate": 1.8972609554746438e-05,
+      "loss": 0.8415,
+      "step": 1683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.864844220133073,
+      "learning_rate": 1.8971154563833624e-05,
+      "loss": 0.7533,
+      "step": 1684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6854041586824038,
+      "learning_rate": 1.896969859924227e-05,
+      "loss": 0.7676,
+      "step": 1685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5650090207778384,
+      "learning_rate": 1.896824166113041e-05,
+      "loss": 0.8221,
+      "step": 1686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7722938808951092,
+      "learning_rate": 1.8966783749656162e-05,
+      "loss": 0.6842,
+      "step": 1687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4914344699971838,
+      "learning_rate": 1.8965324864977766e-05,
+      "loss": 0.6963,
+      "step": 1688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5272008983944905,
+      "learning_rate": 1.8963865007253557e-05,
+      "loss": 0.8026,
+      "step": 1689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.482223785650939,
+      "learning_rate": 1.8962404176641976e-05,
+      "loss": 0.8048,
+      "step": 1690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5593816026928582,
+      "learning_rate": 1.8960942373301584e-05,
+      "loss": 0.6894,
+      "step": 1691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5617523696611526,
+      "learning_rate": 1.895947959739103e-05,
+      "loss": 0.8203,
+      "step": 1692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.814077619268731,
+      "learning_rate": 1.8958015849069074e-05,
+      "loss": 0.8205,
+      "step": 1693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6045784749385548,
+      "learning_rate": 1.8956551128494583e-05,
+      "loss": 0.7982,
+      "step": 1694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7260765769995472,
+      "learning_rate": 1.8955085435826535e-05,
+      "loss": 0.6918,
+      "step": 1695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5504805127465078,
+      "learning_rate": 1.8953618771224003e-05,
+      "loss": 0.7217,
+      "step": 1696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.720269284682822,
+      "learning_rate": 1.895215113484618e-05,
+      "loss": 0.9367,
+      "step": 1697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5714592051560299,
+      "learning_rate": 1.895068252685234e-05,
+      "loss": 0.7853,
+      "step": 1698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7019456115766567,
+      "learning_rate": 1.8949212947401884e-05,
+      "loss": 0.8011,
+      "step": 1699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.884939067526523,
+      "learning_rate": 1.8947742396654318e-05,
+      "loss": 0.8446,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6279141034164122,
+      "learning_rate": 1.8946270874769234e-05,
+      "loss": 0.6527,
+      "step": 1701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.530506408713068,
+      "learning_rate": 1.8944798381906356e-05,
+      "loss": 0.7613,
+      "step": 1702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6168212969929507,
+      "learning_rate": 1.8943324918225495e-05,
+      "loss": 0.7585,
+      "step": 1703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6534578465353755,
+      "learning_rate": 1.894185048388657e-05,
+      "loss": 0.7905,
+      "step": 1704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5523776193348853,
+      "learning_rate": 1.894037507904961e-05,
+      "loss": 0.8521,
+      "step": 1705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.551937987120993,
+      "learning_rate": 1.8938898703874747e-05,
+      "loss": 0.8972,
+      "step": 1706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.69364613713285,
+      "learning_rate": 1.893742135852222e-05,
+      "loss": 0.767,
+      "step": 1707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6210580238903587,
+      "learning_rate": 1.893594304315237e-05,
+      "loss": 0.6865,
+      "step": 1708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.451273066156775,
+      "learning_rate": 1.8934463757925642e-05,
+      "loss": 0.7857,
+      "step": 1709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.683278709031471,
+      "learning_rate": 1.8932983503002598e-05,
+      "loss": 0.7911,
+      "step": 1710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5290904299097812,
+      "learning_rate": 1.8931502278543887e-05,
+      "loss": 0.6458,
+      "step": 1711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5786586347808211,
+      "learning_rate": 1.8930020084710276e-05,
+      "loss": 0.727,
+      "step": 1712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4938119669135754,
+      "learning_rate": 1.8928536921662637e-05,
+      "loss": 0.6879,
+      "step": 1713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5091020987453612,
+      "learning_rate": 1.892705278956194e-05,
+      "loss": 0.6316,
+      "step": 1714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5694048656046364,
+      "learning_rate": 1.892556768856927e-05,
+      "loss": 0.8505,
+      "step": 1715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4532551755755019,
+      "learning_rate": 1.8924081618845804e-05,
+      "loss": 0.6874,
+      "step": 1716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4379667556058273,
+      "learning_rate": 1.892259458055284e-05,
+      "loss": 0.6783,
+      "step": 1717
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5265403642367932,
+      "learning_rate": 1.892110657385176e-05,
+      "loss": 0.7493,
+      "step": 1718
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5479382570256124,
+      "learning_rate": 1.891961759890408e-05,
+      "loss": 0.7337,
+      "step": 1719
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.559889119676712,
+      "learning_rate": 1.891812765587139e-05,
+      "loss": 0.8697,
+      "step": 1720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.593218655173124,
+      "learning_rate": 1.891663674491541e-05,
+      "loss": 0.7259,
+      "step": 1721
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4592818229059854,
+      "learning_rate": 1.8915144866197954e-05,
+      "loss": 0.7734,
+      "step": 1722
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.648324646449686,
+      "learning_rate": 1.8913652019880938e-05,
+      "loss": 0.8242,
+      "step": 1723
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4729500894677385,
+      "learning_rate": 1.8912158206126386e-05,
+      "loss": 0.6893,
+      "step": 1724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6400655969345412,
+      "learning_rate": 1.8910663425096435e-05,
+      "loss": 0.8344,
+      "step": 1725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.497654805250299,
+      "learning_rate": 1.8909167676953317e-05,
+      "loss": 0.7566,
+      "step": 1726
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6871982575991455,
+      "learning_rate": 1.890767096185937e-05,
+      "loss": 0.7531,
+      "step": 1727
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5138489141587188,
+      "learning_rate": 1.8906173279977042e-05,
+      "loss": 0.6804,
+      "step": 1728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6122397020645634,
+      "learning_rate": 1.890467463146888e-05,
+      "loss": 0.804,
+      "step": 1729
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6764245244750045,
+      "learning_rate": 1.890317501649754e-05,
+      "loss": 0.8541,
+      "step": 1730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5251612599991828,
+      "learning_rate": 1.8901674435225784e-05,
+      "loss": 0.7366,
+      "step": 1731
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.719713507717278,
+      "learning_rate": 1.890017288781647e-05,
+      "loss": 0.7156,
+      "step": 1732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5326149533609268,
+      "learning_rate": 1.889867037443258e-05,
+      "loss": 0.8467,
+      "step": 1733
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.686840873303595,
+      "learning_rate": 1.8897166895237172e-05,
+      "loss": 0.7844,
+      "step": 1734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6371997798982931,
+      "learning_rate": 1.889566245039344e-05,
+      "loss": 0.7474,
+      "step": 1735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.611010263494007,
+      "learning_rate": 1.8894157040064657e-05,
+      "loss": 0.6348,
+      "step": 1736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6232898835545204,
+      "learning_rate": 1.889265066441422e-05,
+      "loss": 0.7371,
+      "step": 1737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6149901594670062,
+      "learning_rate": 1.8891143323605618e-05,
+      "loss": 0.7951,
+      "step": 1738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4939178979273533,
+      "learning_rate": 1.888963501780245e-05,
+      "loss": 0.7611,
+      "step": 1739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.569457200983501,
+      "learning_rate": 1.8888125747168418e-05,
+      "loss": 0.7955,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7213407027487273,
+      "learning_rate": 1.8886615511867334e-05,
+      "loss": 0.607,
+      "step": 1741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5760183368741416,
+      "learning_rate": 1.88851043120631e-05,
+      "loss": 0.8016,
+      "step": 1742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7611761834273227,
+      "learning_rate": 1.888359214791975e-05,
+      "loss": 0.8343,
+      "step": 1743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5535745253840973,
+      "learning_rate": 1.888207901960139e-05,
+      "loss": 0.7834,
+      "step": 1744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6699865529676274,
+      "learning_rate": 1.8880564927272255e-05,
+      "loss": 0.7145,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5987541438124577,
+      "learning_rate": 1.887904987109667e-05,
+      "loss": 0.8974,
+      "step": 1746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5848509287747299,
+      "learning_rate": 1.887753385123908e-05,
+      "loss": 0.8937,
+      "step": 1747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5965473543295567,
+      "learning_rate": 1.8876016867864012e-05,
+      "loss": 0.643,
+      "step": 1748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.474355510667469,
+      "learning_rate": 1.8874498921136128e-05,
+      "loss": 0.7083,
+      "step": 1749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6473044578139473,
+      "learning_rate": 1.887298001122016e-05,
+      "loss": 0.8028,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6837953191941109,
+      "learning_rate": 1.8871460138280972e-05,
+      "loss": 0.6944,
+      "step": 1751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6999399008927014,
+      "learning_rate": 1.8869939302483523e-05,
+      "loss": 0.7527,
+      "step": 1752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.580640208318216,
+      "learning_rate": 1.8868417503992868e-05,
+      "loss": 0.7821,
+      "step": 1753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7608845706069354,
+      "learning_rate": 1.8866894742974185e-05,
+      "loss": 0.735,
+      "step": 1754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.783871397792197,
+      "learning_rate": 1.8865371019592734e-05,
+      "loss": 0.6882,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.106234520745424,
+      "learning_rate": 1.8863846334013903e-05,
+      "loss": 0.9047,
+      "step": 1756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6261963712487317,
+      "learning_rate": 1.8862320686403163e-05,
+      "loss": 0.8233,
+      "step": 1757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5592849660065935,
+      "learning_rate": 1.886079407692611e-05,
+      "loss": 0.7271,
+      "step": 1758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.696634596163249,
+      "learning_rate": 1.8859266505748422e-05,
+      "loss": 0.8247,
+      "step": 1759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5583376186257532,
+      "learning_rate": 1.88577379730359e-05,
+      "loss": 0.753,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6329960586391725,
+      "learning_rate": 1.885620847895444e-05,
+      "loss": 0.8578,
+      "step": 1761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.669480616922605,
+      "learning_rate": 1.885467802367005e-05,
+      "loss": 0.7505,
+      "step": 1762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.657482857375443,
+      "learning_rate": 1.8853146607348824e-05,
+      "loss": 0.769,
+      "step": 1763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5154300129562621,
+      "learning_rate": 1.8851614230156984e-05,
+      "loss": 0.749,
+      "step": 1764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6318632109908695,
+      "learning_rate": 1.8850080892260844e-05,
+      "loss": 0.7529,
+      "step": 1765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7001848525559118,
+      "learning_rate": 1.8848546593826827e-05,
+      "loss": 0.8489,
+      "step": 1766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.589134711427665,
+      "learning_rate": 1.8847011335021447e-05,
+      "loss": 0.848,
+      "step": 1767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5385234932969856,
+      "learning_rate": 1.884547511601134e-05,
+      "loss": 0.7197,
+      "step": 1768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5689035447257587,
+      "learning_rate": 1.884393793696324e-05,
+      "loss": 0.7098,
+      "step": 1769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6426536233783764,
+      "learning_rate": 1.884239979804398e-05,
+      "loss": 0.7266,
+      "step": 1770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.677291392580847,
+      "learning_rate": 1.8840860699420497e-05,
+      "loss": 0.8098,
+      "step": 1771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7249955607522947,
+      "learning_rate": 1.8839320641259844e-05,
+      "loss": 0.8916,
+      "step": 1772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6209383073724422,
+      "learning_rate": 1.8837779623729167e-05,
+      "loss": 0.8117,
+      "step": 1773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.593803315763351,
+      "learning_rate": 1.8836237646995717e-05,
+      "loss": 0.8198,
+      "step": 1774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6593476747026599,
+      "learning_rate": 1.883469471122686e-05,
+      "loss": 0.721,
+      "step": 1775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5354092753657136,
+      "learning_rate": 1.8833150816590042e-05,
+      "loss": 0.774,
+      "step": 1776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4254979394593619,
+      "learning_rate": 1.8831605963252845e-05,
+      "loss": 0.6664,
+      "step": 1777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7161205041025949,
+      "learning_rate": 1.883006015138293e-05,
+      "loss": 0.8563,
+      "step": 1778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7088321727886966,
+      "learning_rate": 1.8828513381148072e-05,
+      "loss": 0.8043,
+      "step": 1779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5486816559944219,
+      "learning_rate": 1.8826965652716148e-05,
+      "loss": 0.7855,
+      "step": 1780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3385051756675685,
+      "learning_rate": 1.882541696625514e-05,
+      "loss": 0.7483,
+      "step": 1781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.498188058539235,
+      "learning_rate": 1.8823867321933135e-05,
+      "loss": 0.7488,
+      "step": 1782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.736701746116013,
+      "learning_rate": 1.882231671991832e-05,
+      "loss": 0.805,
+      "step": 1783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7067130208010808,
+      "learning_rate": 1.8820765160378996e-05,
+      "loss": 0.7416,
+      "step": 1784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5944026426589217,
+      "learning_rate": 1.881921264348355e-05,
+      "loss": 0.863,
+      "step": 1785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.63057640894758,
+      "learning_rate": 1.8817659169400493e-05,
+      "loss": 0.8167,
+      "step": 1786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4701737654650993,
+      "learning_rate": 1.8816104738298425e-05,
+      "loss": 0.7732,
+      "step": 1787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7659716645257424,
+      "learning_rate": 1.8814549350346052e-05,
+      "loss": 0.8413,
+      "step": 1788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6703190187917394,
+      "learning_rate": 1.8812993005712196e-05,
+      "loss": 0.8808,
+      "step": 1789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5804918076804337,
+      "learning_rate": 1.881143570456577e-05,
+      "loss": 0.8396,
+      "step": 1790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.502015060915155,
+      "learning_rate": 1.880987744707579e-05,
+      "loss": 0.7767,
+      "step": 1791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5100557185000367,
+      "learning_rate": 1.8808318233411384e-05,
+      "loss": 0.837,
+      "step": 1792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4980699724605326,
+      "learning_rate": 1.880675806374178e-05,
+      "loss": 0.8265,
+      "step": 1793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6145024202929508,
+      "learning_rate": 1.880519693823631e-05,
+      "loss": 0.8522,
+      "step": 1794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7060963762719894,
+      "learning_rate": 1.8803634857064413e-05,
+      "loss": 0.6858,
+      "step": 1795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.66699583908313,
+      "learning_rate": 1.8802071820395626e-05,
+      "loss": 0.8265,
+      "step": 1796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5088819587635958,
+      "learning_rate": 1.880050782839959e-05,
+      "loss": 0.8256,
+      "step": 1797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8248098806211075,
+      "learning_rate": 1.8798942881246057e-05,
+      "loss": 0.7441,
+      "step": 1798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.497823717757942,
+      "learning_rate": 1.8797376979104874e-05,
+      "loss": 0.6212,
+      "step": 1799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5313709881390287,
+      "learning_rate": 1.879581012214599e-05,
+      "loss": 0.7585,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6533852967630702,
+      "learning_rate": 1.8794242310539475e-05,
+      "loss": 0.706,
+      "step": 1801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.57322286965956,
+      "learning_rate": 1.879267354445548e-05,
+      "loss": 0.7315,
+      "step": 1802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4688226723378812,
+      "learning_rate": 1.8791103824064275e-05,
+      "loss": 0.6824,
+      "step": 1803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6606304786948904,
+      "learning_rate": 1.8789533149536227e-05,
+      "loss": 0.774,
+      "step": 1804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4170226955215965,
+      "learning_rate": 1.8787961521041808e-05,
+      "loss": 0.7616,
+      "step": 1805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6319528691892806,
+      "learning_rate": 1.8786388938751593e-05,
+      "loss": 0.7277,
+      "step": 1806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7368883126240444,
+      "learning_rate": 1.8784815402836264e-05,
+      "loss": 0.7857,
+      "step": 1807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6956446906840685,
+      "learning_rate": 1.87832409134666e-05,
+      "loss": 0.8243,
+      "step": 1808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5661135765652143,
+      "learning_rate": 1.8781665470813493e-05,
+      "loss": 0.7309,
+      "step": 1809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6488344290849406,
+      "learning_rate": 1.878008907504793e-05,
+      "loss": 0.741,
+      "step": 1810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.748004174675965,
+      "learning_rate": 1.8778511726341e-05,
+      "loss": 0.812,
+      "step": 1811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6500412921236254,
+      "learning_rate": 1.8776933424863903e-05,
+      "loss": 0.75,
+      "step": 1812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5783811495293991,
+      "learning_rate": 1.8775354170787938e-05,
+      "loss": 0.7833,
+      "step": 1813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6438030612284624,
+      "learning_rate": 1.8773773964284512e-05,
+      "loss": 0.8352,
+      "step": 1814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7254538324598188,
+      "learning_rate": 1.8772192805525125e-05,
+      "loss": 0.7913,
+      "step": 1815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4609001559387844,
+      "learning_rate": 1.8770610694681393e-05,
+      "loss": 0.7164,
+      "step": 1816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5638189485710678,
+      "learning_rate": 1.8769027631925027e-05,
+      "loss": 0.741,
+      "step": 1817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5970980794499663,
+      "learning_rate": 1.876744361742784e-05,
+      "loss": 0.752,
+      "step": 1818
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3691889410857434,
+      "learning_rate": 1.876585865136176e-05,
+      "loss": 0.7374,
+      "step": 1819
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7025448759231918,
+      "learning_rate": 1.8764272733898808e-05,
+      "loss": 0.7367,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6286596973353795,
+      "learning_rate": 1.8762685865211106e-05,
+      "loss": 0.7936,
+      "step": 1821
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8002277034115608,
+      "learning_rate": 1.8761098045470887e-05,
+      "loss": 0.8227,
+      "step": 1822
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6901918456825955,
+      "learning_rate": 1.875950927485048e-05,
+      "loss": 0.7645,
+      "step": 1823
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5603243747675015,
+      "learning_rate": 1.875791955352233e-05,
+      "loss": 0.7471,
+      "step": 1824
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.638513369552182,
+      "learning_rate": 1.8756328881658968e-05,
+      "loss": 0.86,
+      "step": 1825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4728899465231686,
+      "learning_rate": 1.875473725943304e-05,
+      "loss": 0.7207,
+      "step": 1826
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6494863709917857,
+      "learning_rate": 1.8753144687017292e-05,
+      "loss": 0.755,
+      "step": 1827
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4154536873183075,
+      "learning_rate": 1.8751551164584568e-05,
+      "loss": 0.7739,
+      "step": 1828
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.58099440095728,
+      "learning_rate": 1.8749956692307825e-05,
+      "loss": 0.7177,
+      "step": 1829
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.526424937636552,
+      "learning_rate": 1.874836127036012e-05,
+      "loss": 0.6252,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7385896248430377,
+      "learning_rate": 1.874676489891461e-05,
+      "loss": 0.75,
+      "step": 1831
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6335448597236224,
+      "learning_rate": 1.8745167578144552e-05,
+      "loss": 0.8474,
+      "step": 1832
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5534145226465848,
+      "learning_rate": 1.8743569308223312e-05,
+      "loss": 0.7913,
+      "step": 1833
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5581638165153364,
+      "learning_rate": 1.8741970089324357e-05,
+      "loss": 0.6453,
+      "step": 1834
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6011044528243155,
+      "learning_rate": 1.874036992162126e-05,
+      "loss": 0.7914,
+      "step": 1835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.768395282832198,
+      "learning_rate": 1.873876880528769e-05,
+      "loss": 0.7309,
+      "step": 1836
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5493951226784552,
+      "learning_rate": 1.8737166740497427e-05,
+      "loss": 0.7553,
+      "step": 1837
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5239649939874127,
+      "learning_rate": 1.873556372742435e-05,
+      "loss": 0.7589,
+      "step": 1838
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.634050189294019,
+      "learning_rate": 1.8733959766242435e-05,
+      "loss": 0.897,
+      "step": 1839
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6207931351883975,
+      "learning_rate": 1.8732354857125773e-05,
+      "loss": 0.814,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.507728990025924,
+      "learning_rate": 1.873074900024855e-05,
+      "loss": 0.6828,
+      "step": 1841
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5626504219753878,
+      "learning_rate": 1.8729142195785057e-05,
+      "loss": 0.8172,
+      "step": 1842
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6739510888151425,
+      "learning_rate": 1.8727534443909686e-05,
+      "loss": 0.7512,
+      "step": 1843
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5598058732542022,
+      "learning_rate": 1.872592574479694e-05,
+      "loss": 0.7194,
+      "step": 1844
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6005303628723884,
+      "learning_rate": 1.8724316098621405e-05,
+      "loss": 0.7972,
+      "step": 1845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.600654059697692,
+      "learning_rate": 1.872270550555779e-05,
+      "loss": 0.9066,
+      "step": 1846
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5518927935254088,
+      "learning_rate": 1.872109396578091e-05,
+      "loss": 0.6793,
+      "step": 1847
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5799506697248105,
+      "learning_rate": 1.8719481479465657e-05,
+      "loss": 0.6856,
+      "step": 1848
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6159705842129906,
+      "learning_rate": 1.8717868046787046e-05,
+      "loss": 0.8038,
+      "step": 1849
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4684918042317172,
+      "learning_rate": 1.871625366792019e-05,
+      "loss": 0.6801,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6412423287402154,
+      "learning_rate": 1.8714638343040306e-05,
+      "loss": 0.7523,
+      "step": 1851
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6437306891728323,
+      "learning_rate": 1.871302207232271e-05,
+      "loss": 0.7218,
+      "step": 1852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4601714173931502,
+      "learning_rate": 1.871140485594283e-05,
+      "loss": 0.7227,
+      "step": 1853
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6513526696248817,
+      "learning_rate": 1.8709786694076178e-05,
+      "loss": 0.8656,
+      "step": 1854
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4853149556610323,
+      "learning_rate": 1.870816758689839e-05,
+      "loss": 0.7308,
+      "step": 1855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6802466029990593,
+      "learning_rate": 1.870654753458519e-05,
+      "loss": 0.8212,
+      "step": 1856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8500457105225216,
+      "learning_rate": 1.8704926537312408e-05,
+      "loss": 0.8009,
+      "step": 1857
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5972154532993916,
+      "learning_rate": 1.8703304595255983e-05,
+      "loss": 0.6746,
+      "step": 1858
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6662701350599647,
+      "learning_rate": 1.8701681708591947e-05,
+      "loss": 0.8041,
+      "step": 1859
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5730426403183335,
+      "learning_rate": 1.8700057877496443e-05,
+      "loss": 0.9202,
+      "step": 1860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4563945916902619,
+      "learning_rate": 1.8698433102145706e-05,
+      "loss": 0.7369,
+      "step": 1861
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4702139366238451,
+      "learning_rate": 1.8696807382716085e-05,
+      "loss": 0.7116,
+      "step": 1862
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5621107416457056,
+      "learning_rate": 1.869518071938403e-05,
+      "loss": 0.7758,
+      "step": 1863
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5895626444726687,
+      "learning_rate": 1.8693553112326084e-05,
+      "loss": 0.8106,
+      "step": 1864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7607266030575186,
+      "learning_rate": 1.8691924561718897e-05,
+      "loss": 0.844,
+      "step": 1865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4933064378757956,
+      "learning_rate": 1.8690295067739226e-05,
+      "loss": 0.7138,
+      "step": 1866
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.712265806011729,
+      "learning_rate": 1.8688664630563928e-05,
+      "loss": 0.8637,
+      "step": 1867
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6019635887840478,
+      "learning_rate": 1.8687033250369955e-05,
+      "loss": 0.7653,
+      "step": 1868
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.612276802897441,
+      "learning_rate": 1.8685400927334377e-05,
+      "loss": 0.8336,
+      "step": 1869
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.670180632071749,
+      "learning_rate": 1.8683767661634354e-05,
+      "loss": 0.8118,
+      "step": 1870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3743265251696855,
+      "learning_rate": 1.8682133453447147e-05,
+      "loss": 0.82,
+      "step": 1871
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.683875985160119,
+      "learning_rate": 1.868049830295013e-05,
+      "loss": 0.9025,
+      "step": 1872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6648127443357852,
+      "learning_rate": 1.867886221032077e-05,
+      "loss": 0.8204,
+      "step": 1873
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6458275270596272,
+      "learning_rate": 1.8677225175736636e-05,
+      "loss": 0.7923,
+      "step": 1874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5987414009006289,
+      "learning_rate": 1.8675587199375407e-05,
+      "loss": 0.6926,
+      "step": 1875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.801676353452884,
+      "learning_rate": 1.8673948281414857e-05,
+      "loss": 0.8069,
+      "step": 1876
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5174117497935236,
+      "learning_rate": 1.867230842203287e-05,
+      "loss": 0.6795,
+      "step": 1877
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.447289592777005,
+      "learning_rate": 1.8670667621407423e-05,
+      "loss": 0.7696,
+      "step": 1878
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.543875258571506,
+      "learning_rate": 1.8669025879716597e-05,
+      "loss": 0.7519,
+      "step": 1879
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6415775747071424,
+      "learning_rate": 1.866738319713858e-05,
+      "loss": 0.7268,
+      "step": 1880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7785108692805012,
+      "learning_rate": 1.8665739573851662e-05,
+      "loss": 0.7699,
+      "step": 1881
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7857674261403411,
+      "learning_rate": 1.866409501003423e-05,
+      "loss": 0.8229,
+      "step": 1882
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9847860216254964,
+      "learning_rate": 1.8662449505864776e-05,
+      "loss": 0.7704,
+      "step": 1883
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7227834593591063,
+      "learning_rate": 1.8660803061521894e-05,
+      "loss": 0.7932,
+      "step": 1884
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6255355103927502,
+      "learning_rate": 1.865915567718428e-05,
+      "loss": 0.682,
+      "step": 1885
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5679640202331104,
+      "learning_rate": 1.865750735303073e-05,
+      "loss": 0.812,
+      "step": 1886
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5657907851284936,
+      "learning_rate": 1.8655858089240143e-05,
+      "loss": 0.7679,
+      "step": 1887
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3940385718601986,
+      "learning_rate": 1.8654207885991527e-05,
+      "loss": 0.6367,
+      "step": 1888
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.570281967703175,
+      "learning_rate": 1.865255674346398e-05,
+      "loss": 0.7764,
+      "step": 1889
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6446730773550722,
+      "learning_rate": 1.8650904661836707e-05,
+      "loss": 0.7509,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7466286168189615,
+      "learning_rate": 1.864925164128902e-05,
+      "loss": 0.7999,
+      "step": 1891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5994431575638575,
+      "learning_rate": 1.864759768200033e-05,
+      "loss": 0.7346,
+      "step": 1892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6980032406179928,
+      "learning_rate": 1.864594278415014e-05,
+      "loss": 0.8921,
+      "step": 1893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6641202788730192,
+      "learning_rate": 1.864428694791807e-05,
+      "loss": 0.7413,
+      "step": 1894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4534612756573948,
+      "learning_rate": 1.8642630173483832e-05,
+      "loss": 0.7873,
+      "step": 1895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6806786748699862,
+      "learning_rate": 1.8640972461027246e-05,
+      "loss": 0.6926,
+      "step": 1896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7567742933825417,
+      "learning_rate": 1.863931381072823e-05,
+      "loss": 0.806,
+      "step": 1897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5208842194437813,
+      "learning_rate": 1.8637654222766802e-05,
+      "loss": 0.7438,
+      "step": 1898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5697337897387504,
+      "learning_rate": 1.8635993697323086e-05,
+      "loss": 0.8382,
+      "step": 1899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5281416666058911,
+      "learning_rate": 1.8634332234577307e-05,
+      "loss": 0.7687,
+      "step": 1900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5959906743154828,
+      "learning_rate": 1.8632669834709787e-05,
+      "loss": 0.7747,
+      "step": 1901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.553663965084361,
+      "learning_rate": 1.8631006497900957e-05,
+      "loss": 0.7009,
+      "step": 1902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5721802721916782,
+      "learning_rate": 1.862934222433135e-05,
+      "loss": 0.7484,
+      "step": 1903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.523293199579708,
+      "learning_rate": 1.8627677014181586e-05,
+      "loss": 0.6594,
+      "step": 1904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.615424841576138,
+      "learning_rate": 1.862601086763241e-05,
+      "loss": 0.7287,
+      "step": 1905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4620235693583123,
+      "learning_rate": 1.8624343784864644e-05,
+      "loss": 0.6631,
+      "step": 1906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.609922327362643,
+      "learning_rate": 1.8622675766059232e-05,
+      "loss": 0.6958,
+      "step": 1907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7415547413413075,
+      "learning_rate": 1.862100681139721e-05,
+      "loss": 0.8045,
+      "step": 1908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.642979877233161,
+      "learning_rate": 1.861933692105972e-05,
+      "loss": 0.8172,
+      "step": 1909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5667632726209306,
+      "learning_rate": 1.8617666095227994e-05,
+      "loss": 0.7247,
+      "step": 1910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4583686501782427,
+      "learning_rate": 1.8615994334083377e-05,
+      "loss": 0.81,
+      "step": 1911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4396267855158282,
+      "learning_rate": 1.8614321637807315e-05,
+      "loss": 0.7329,
+      "step": 1912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6642235119219537,
+      "learning_rate": 1.8612648006581354e-05,
+      "loss": 0.8072,
+      "step": 1913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.754213812581633,
+      "learning_rate": 1.861097344058714e-05,
+      "loss": 0.8149,
+      "step": 1914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6761790132590595,
+      "learning_rate": 1.8609297940006418e-05,
+      "loss": 0.8206,
+      "step": 1915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.560191341344239,
+      "learning_rate": 1.8607621505021035e-05,
+      "loss": 0.7154,
+      "step": 1916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4918257116316898,
+      "learning_rate": 1.860594413581295e-05,
+      "loss": 0.7098,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5958708627476872,
+      "learning_rate": 1.860426583256421e-05,
+      "loss": 0.8249,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4247394782586766,
+      "learning_rate": 1.8602586595456974e-05,
+      "loss": 0.8818,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7522243225471708,
+      "learning_rate": 1.8600906424673487e-05,
+      "loss": 0.8321,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5649963909661708,
+      "learning_rate": 1.8599225320396113e-05,
+      "loss": 0.7594,
+      "step": 1921
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6616037284068703,
+      "learning_rate": 1.8597543282807303e-05,
+      "loss": 0.808,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6696455666594916,
+      "learning_rate": 1.8595860312089625e-05,
+      "loss": 0.7538,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6723167707677005,
+      "learning_rate": 1.859417640842573e-05,
+      "loss": 0.768,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6174241472922297,
+      "learning_rate": 1.859249157199839e-05,
+      "loss": 0.7958,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4890001345357784,
+      "learning_rate": 1.8590805802990458e-05,
+      "loss": 0.797,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5636449533652899,
+      "learning_rate": 1.8589119101584902e-05,
+      "loss": 0.7467,
+      "step": 1927
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7449087745880458,
+      "learning_rate": 1.8587431467964784e-05,
+      "loss": 0.8519,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6217747543112395,
+      "learning_rate": 1.8585742902313274e-05,
+      "loss": 0.7736,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5118301003504073,
+      "learning_rate": 1.858405340481364e-05,
+      "loss": 0.6824,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4925272292019434,
+      "learning_rate": 1.8582362975649245e-05,
+      "loss": 0.678,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6573874004589289,
+      "learning_rate": 1.8580671615003566e-05,
+      "loss": 0.8158,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5882708936318732,
+      "learning_rate": 1.8578979323060164e-05,
+      "loss": 0.703,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.747937980589991,
+      "learning_rate": 1.8577286100002723e-05,
+      "loss": 0.8432,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5880321948553646,
+      "learning_rate": 1.8575591946015006e-05,
+      "loss": 0.8102,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.67263280842425,
+      "learning_rate": 1.8573896861280893e-05,
+      "loss": 0.8489,
+      "step": 1936
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.616183957035531,
+      "learning_rate": 1.8572200845984352e-05,
+      "loss": 0.6341,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6906920012154358,
+      "learning_rate": 1.857050390030947e-05,
+      "loss": 0.8366,
+      "step": 1938
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5695251712587637,
+      "learning_rate": 1.8568806024440415e-05,
+      "loss": 0.6616,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.531032808164511,
+      "learning_rate": 1.8567107218561463e-05,
+      "loss": 0.7847,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5605863172120555,
+      "learning_rate": 1.8565407482857e-05,
+      "loss": 0.7061,
+      "step": 1941
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6497185295835817,
+      "learning_rate": 1.85637068175115e-05,
+      "loss": 0.7621,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6176729022633565,
+      "learning_rate": 1.856200522270955e-05,
+      "loss": 0.7856,
+      "step": 1943
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9537736544492905,
+      "learning_rate": 1.856030269863583e-05,
+      "loss": 0.88,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5430187767275976,
+      "learning_rate": 1.8558599245475115e-05,
+      "loss": 0.6248,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3672398031794841,
+      "learning_rate": 1.8556894863412297e-05,
+      "loss": 0.6678,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6851967538360875,
+      "learning_rate": 1.8555189552632353e-05,
+      "loss": 0.7628,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.308925529168711,
+      "learning_rate": 1.8553483313320372e-05,
+      "loss": 0.7125,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5482666375476917,
+      "learning_rate": 1.855177614566154e-05,
+      "loss": 0.7105,
+      "step": 1949
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6451222261319363,
+      "learning_rate": 1.8550068049841143e-05,
+      "loss": 0.754,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.530070822063962,
+      "learning_rate": 1.8548359026044567e-05,
+      "loss": 0.8445,
+      "step": 1951
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6759016729134624,
+      "learning_rate": 1.85466490744573e-05,
+      "loss": 0.8776,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6564499377083384,
+      "learning_rate": 1.854493819526493e-05,
+      "loss": 0.8513,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6158099814548028,
+      "learning_rate": 1.854322638865315e-05,
+      "loss": 0.8304,
+      "step": 1954
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6128948769079847,
+      "learning_rate": 1.854151365480774e-05,
+      "loss": 0.7835,
+      "step": 1955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6422859708402293,
+      "learning_rate": 1.8539799993914602e-05,
+      "loss": 0.7945,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5442670681243345,
+      "learning_rate": 1.8538085406159722e-05,
+      "loss": 0.7909,
+      "step": 1957
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6477871595113218,
+      "learning_rate": 1.8536369891729188e-05,
+      "loss": 0.7269,
+      "step": 1958
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6111243464475271,
+      "learning_rate": 1.85346534508092e-05,
+      "loss": 0.7188,
+      "step": 1959
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.528940302181452,
+      "learning_rate": 1.8532936083586047e-05,
+      "loss": 0.757,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6760006322163905,
+      "learning_rate": 1.853121779024612e-05,
+      "loss": 0.6481,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5850301760975005,
+      "learning_rate": 1.8529498570975918e-05,
+      "loss": 0.8071,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.439301043406329,
+      "learning_rate": 1.852777842596203e-05,
+      "loss": 0.6926,
+      "step": 1963
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4913418883840095,
+      "learning_rate": 1.8526057355391153e-05,
+      "loss": 0.7353,
+      "step": 1964
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4632243798516378,
+      "learning_rate": 1.8524335359450084e-05,
+      "loss": 0.6683,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6979225750276388,
+      "learning_rate": 1.852261243832572e-05,
+      "loss": 0.7964,
+      "step": 1966
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6935771936979542,
+      "learning_rate": 1.852088859220505e-05,
+      "loss": 0.8466,
+      "step": 1967
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6235813178922434,
+      "learning_rate": 1.851916382127518e-05,
+      "loss": 0.7163,
+      "step": 1968
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6952450683272433,
+      "learning_rate": 1.85174381257233e-05,
+      "loss": 0.8911,
+      "step": 1969
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6217638700128432,
+      "learning_rate": 1.8515711505736708e-05,
+      "loss": 0.8231,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5461320142762094,
+      "learning_rate": 1.8513983961502802e-05,
+      "loss": 0.6862,
+      "step": 1971
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8345041401665254,
+      "learning_rate": 1.851225549320908e-05,
+      "loss": 0.7441,
+      "step": 1972
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5711831067423128,
+      "learning_rate": 1.8510526101043146e-05,
+      "loss": 0.869,
+      "step": 1973
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5191116397427378,
+      "learning_rate": 1.850879578519269e-05,
+      "loss": 0.7696,
+      "step": 1974
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5090778444514343,
+      "learning_rate": 1.8507064545845513e-05,
+      "loss": 0.7709,
+      "step": 1975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5043235651481828,
+      "learning_rate": 1.8505332383189518e-05,
+      "loss": 0.6482,
+      "step": 1976
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7494370121789027,
+      "learning_rate": 1.85035992974127e-05,
+      "loss": 0.659,
+      "step": 1977
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6977092603202606,
+      "learning_rate": 1.850186528870316e-05,
+      "loss": 0.8217,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4858417843522007,
+      "learning_rate": 1.85001303572491e-05,
+      "loss": 0.7733,
+      "step": 1979
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3671668705036,
+      "learning_rate": 1.8498394503238814e-05,
+      "loss": 0.7278,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6825392973325513,
+      "learning_rate": 1.84966577268607e-05,
+      "loss": 0.7464,
+      "step": 1981
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.555401729857369,
+      "learning_rate": 1.849492002830327e-05,
+      "loss": 0.766,
+      "step": 1982
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6489193347651052,
+      "learning_rate": 1.8493181407755117e-05,
+      "loss": 0.8089,
+      "step": 1983
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.555459995400608,
+      "learning_rate": 1.849144186540494e-05,
+      "loss": 0.7948,
+      "step": 1984
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6696539094363,
+      "learning_rate": 1.8489701401441534e-05,
+      "loss": 0.7851,
+      "step": 1985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.599201865291812,
+      "learning_rate": 1.848796001605381e-05,
+      "loss": 0.8093,
+      "step": 1986
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6274974091557912,
+      "learning_rate": 1.8486217709430757e-05,
+      "loss": 0.7926,
+      "step": 1987
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3623804172978458,
+      "learning_rate": 1.848447448176149e-05,
+      "loss": 0.7262,
+      "step": 1988
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.629211820742282,
+      "learning_rate": 1.8482730333235196e-05,
+      "loss": 0.8294,
+      "step": 1989
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7541029673914632,
+      "learning_rate": 1.8480985264041176e-05,
+      "loss": 0.8159,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.586614609609062,
+      "learning_rate": 1.847923927436884e-05,
+      "loss": 0.6823,
+      "step": 1991
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5419850233142978,
+      "learning_rate": 1.8477492364407677e-05,
+      "loss": 0.6992,
+      "step": 1992
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.472491430080609,
+      "learning_rate": 1.8475744534347293e-05,
+      "loss": 0.7535,
+      "step": 1993
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5645681274063772,
+      "learning_rate": 1.8473995784377384e-05,
+      "loss": 0.7991,
+      "step": 1994
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6389601170363615,
+      "learning_rate": 1.8472246114687754e-05,
+      "loss": 0.7141,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6616292369048578,
+      "learning_rate": 1.8470495525468295e-05,
+      "loss": 0.7879,
+      "step": 1996
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4737315897382532,
+      "learning_rate": 1.8468744016909012e-05,
+      "loss": 0.7577,
+      "step": 1997
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.549146981818495,
+      "learning_rate": 1.8466991589200004e-05,
+      "loss": 0.6605,
+      "step": 1998
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4417487564219365,
+      "learning_rate": 1.8465238242531467e-05,
+      "loss": 0.7124,
+      "step": 1999
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6835586343638251,
+      "learning_rate": 1.84634839770937e-05,
+      "loss": 0.7829,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7115119058166468,
+      "learning_rate": 1.8461728793077104e-05,
+      "loss": 0.7586,
+      "step": 2001
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4947857186720843,
+      "learning_rate": 1.8459972690672172e-05,
+      "loss": 0.7328,
+      "step": 2002
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4280227740430842,
+      "learning_rate": 1.8458215670069502e-05,
+      "loss": 0.7671,
+      "step": 2003
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6151555542402964,
+      "learning_rate": 1.8456457731459795e-05,
+      "loss": 0.848,
+      "step": 2004
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.479691535263727,
+      "learning_rate": 1.8454698875033843e-05,
+      "loss": 0.8455,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4177021763508002,
+      "learning_rate": 1.8452939100982547e-05,
+      "loss": 0.7119,
+      "step": 2006
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6199388078799257,
+      "learning_rate": 1.8451178409496903e-05,
+      "loss": 0.6344,
+      "step": 2007
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5825791681632486,
+      "learning_rate": 1.8449416800767998e-05,
+      "loss": 0.7744,
+      "step": 2008
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5318746508145062,
+      "learning_rate": 1.8447654274987038e-05,
+      "loss": 0.8559,
+      "step": 2009
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6233848458956928,
+      "learning_rate": 1.844589083234531e-05,
+      "loss": 0.7345,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.994553363314384,
+      "learning_rate": 1.8444126473034212e-05,
+      "loss": 0.7971,
+      "step": 2011
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6031955233721904,
+      "learning_rate": 1.844236119724524e-05,
+      "loss": 0.7934,
+      "step": 2012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4580459780767323,
+      "learning_rate": 1.8440595005169985e-05,
+      "loss": 0.7894,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7207833492032534,
+      "learning_rate": 1.843882789700013e-05,
+      "loss": 0.8159,
+      "step": 2014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.538572416456644,
+      "learning_rate": 1.843705987292748e-05,
+      "loss": 0.8205,
+      "step": 2015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6077079023161813,
+      "learning_rate": 1.8435290933143925e-05,
+      "loss": 0.712,
+      "step": 2016
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4914180648119735,
+      "learning_rate": 1.8433521077841447e-05,
+      "loss": 0.705,
+      "step": 2017
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5844200282792549,
+      "learning_rate": 1.8431750307212143e-05,
+      "loss": 0.747,
+      "step": 2018
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5512967626223364,
+      "learning_rate": 1.84299786214482e-05,
+      "loss": 0.7392,
+      "step": 2019
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6445683986761288,
+      "learning_rate": 1.8428206020741913e-05,
+      "loss": 0.8185,
+      "step": 2020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5331188402336782,
+      "learning_rate": 1.8426432505285658e-05,
+      "loss": 0.6476,
+      "step": 2021
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5679496580351822,
+      "learning_rate": 1.8424658075271934e-05,
+      "loss": 0.6425,
+      "step": 2022
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5806644443693592,
+      "learning_rate": 1.8422882730893323e-05,
+      "loss": 0.815,
+      "step": 2023
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8126513267054625,
+      "learning_rate": 1.8421106472342507e-05,
+      "loss": 0.7919,
+      "step": 2024
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8010077865252379,
+      "learning_rate": 1.841932929981228e-05,
+      "loss": 0.792,
+      "step": 2025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6176071870755673,
+      "learning_rate": 1.8417551213495516e-05,
+      "loss": 0.681,
+      "step": 2026
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6194212121101104,
+      "learning_rate": 1.8415772213585206e-05,
+      "loss": 0.781,
+      "step": 2027
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4615424765257328,
+      "learning_rate": 1.8413992300274432e-05,
+      "loss": 0.7386,
+      "step": 2028
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.345972388499806,
+      "learning_rate": 1.8412211473756366e-05,
+      "loss": 0.6836,
+      "step": 2029
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.420925984429223,
+      "learning_rate": 1.8410429734224305e-05,
+      "loss": 0.7123,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4714918647469502,
+      "learning_rate": 1.8408647081871617e-05,
+      "loss": 0.7548,
+      "step": 2031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2880187247176165,
+      "learning_rate": 1.8406863516891787e-05,
+      "loss": 0.6705,
+      "step": 2032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5561890649523946,
+      "learning_rate": 1.840507903947839e-05,
+      "loss": 0.8041,
+      "step": 2033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8240869988638713,
+      "learning_rate": 1.8403293649825105e-05,
+      "loss": 0.8352,
+      "step": 2034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6843780737067835,
+      "learning_rate": 1.8401507348125706e-05,
+      "loss": 0.8211,
+      "step": 2035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8174509816114721,
+      "learning_rate": 1.8399720134574068e-05,
+      "loss": 0.8882,
+      "step": 2036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8972402872295238,
+      "learning_rate": 1.839793200936417e-05,
+      "loss": 0.8591,
+      "step": 2037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7514471330221437,
+      "learning_rate": 1.8396142972690075e-05,
+      "loss": 0.6459,
+      "step": 2038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8239195890199404,
+      "learning_rate": 1.8394353024745965e-05,
+      "loss": 0.7782,
+      "step": 2039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5675086725687128,
+      "learning_rate": 1.839256216572611e-05,
+      "loss": 0.8358,
+      "step": 2040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5085805660216012,
+      "learning_rate": 1.8390770395824874e-05,
+      "loss": 0.735,
+      "step": 2041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.547175912059753,
+      "learning_rate": 1.8388977715236728e-05,
+      "loss": 0.7459,
+      "step": 2042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7356081253311848,
+      "learning_rate": 1.838718412415624e-05,
+      "loss": 0.8296,
+      "step": 2043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.655669317138079,
+      "learning_rate": 1.8385389622778076e-05,
+      "loss": 0.9693,
+      "step": 2044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6362615478232747,
+      "learning_rate": 1.8383594211297002e-05,
+      "loss": 0.8121,
+      "step": 2045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5372469664900423,
+      "learning_rate": 1.838179788990788e-05,
+      "loss": 0.7378,
+      "step": 2046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7855729505540894,
+      "learning_rate": 1.838000065880568e-05,
+      "loss": 0.8149,
+      "step": 2047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6119552242718482,
+      "learning_rate": 1.837820251818545e-05,
+      "loss": 0.7797,
+      "step": 2048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6814490296118239,
+      "learning_rate": 1.837640346824236e-05,
+      "loss": 0.6792,
+      "step": 2049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5907274351368677,
+      "learning_rate": 1.837460350917166e-05,
+      "loss": 0.6379,
+      "step": 2050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.582028706018697,
+      "learning_rate": 1.837280264116872e-05,
+      "loss": 0.87,
+      "step": 2051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5896426368682781,
+      "learning_rate": 1.837100086442899e-05,
+      "loss": 0.7454,
+      "step": 2052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.515693029722567,
+      "learning_rate": 1.8369198179148022e-05,
+      "loss": 0.5782,
+      "step": 2053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6823168519943408,
+      "learning_rate": 1.836739458552147e-05,
+      "loss": 0.7646,
+      "step": 2054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5807384127341129,
+      "learning_rate": 1.8365590083745085e-05,
+      "loss": 0.7864,
+      "step": 2055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5014690904678236,
+      "learning_rate": 1.8363784674014726e-05,
+      "loss": 0.7983,
+      "step": 2056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.527163447694033,
+      "learning_rate": 1.836197835652633e-05,
+      "loss": 0.7369,
+      "step": 2057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4761009440956374,
+      "learning_rate": 1.8360171131475954e-05,
+      "loss": 0.7562,
+      "step": 2058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5807536940605051,
+      "learning_rate": 1.8358362999059738e-05,
+      "loss": 0.7055,
+      "step": 2059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7363606674901761,
+      "learning_rate": 1.835655395947393e-05,
+      "loss": 0.8579,
+      "step": 2060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5892285709310214,
+      "learning_rate": 1.835474401291487e-05,
+      "loss": 0.76,
+      "step": 2061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.96710212615678,
+      "learning_rate": 1.8352933159579e-05,
+      "loss": 0.7318,
+      "step": 2062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.656354040552658,
+      "learning_rate": 1.8351121399662862e-05,
+      "loss": 0.7117,
+      "step": 2063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4360631305910538,
+      "learning_rate": 1.8349308733363093e-05,
+      "loss": 0.6603,
+      "step": 2064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.646228813200624,
+      "learning_rate": 1.8347495160876432e-05,
+      "loss": 0.7216,
+      "step": 2065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.446053013317963,
+      "learning_rate": 1.834568068239971e-05,
+      "loss": 0.7162,
+      "step": 2066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4323835655155246,
+      "learning_rate": 1.8343865298129858e-05,
+      "loss": 0.7047,
+      "step": 2067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5035102946772407,
+      "learning_rate": 1.8342049008263917e-05,
+      "loss": 0.8284,
+      "step": 2068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6444469347806443,
+      "learning_rate": 1.8340231812999007e-05,
+      "loss": 0.7184,
+      "step": 2069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6596641180910732,
+      "learning_rate": 1.8338413712532365e-05,
+      "loss": 0.8087,
+      "step": 2070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5875512346960872,
+      "learning_rate": 1.833659470706131e-05,
+      "loss": 0.7649,
+      "step": 2071
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5843463610866617,
+      "learning_rate": 1.8334774796783268e-05,
+      "loss": 0.7858,
+      "step": 2072
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4945176583565407,
+      "learning_rate": 1.833295398189576e-05,
+      "loss": 0.7108,
+      "step": 2073
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.812069632882091,
+      "learning_rate": 1.8331132262596418e-05,
+      "loss": 0.8371,
+      "step": 2074
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4764660495567634,
+      "learning_rate": 1.832930963908295e-05,
+      "loss": 0.6777,
+      "step": 2075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5515759927728716,
+      "learning_rate": 1.8327486111553174e-05,
+      "loss": 0.8229,
+      "step": 2076
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4658649863388442,
+      "learning_rate": 1.832566168020501e-05,
+      "loss": 0.7533,
+      "step": 2077
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5142633214554027,
+      "learning_rate": 1.832383634523647e-05,
+      "loss": 0.6924,
+      "step": 2078
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5895282587876711,
+      "learning_rate": 1.8322010106845663e-05,
+      "loss": 0.8646,
+      "step": 2079
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4819718119255312,
+      "learning_rate": 1.8320182965230803e-05,
+      "loss": 0.6515,
+      "step": 2080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5681585082866136,
+      "learning_rate": 1.8318354920590195e-05,
+      "loss": 0.8381,
+      "step": 2081
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4810855763092055,
+      "learning_rate": 1.8316525973122243e-05,
+      "loss": 0.769,
+      "step": 2082
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7651158078259417,
+      "learning_rate": 1.8314696123025456e-05,
+      "loss": 0.8944,
+      "step": 2083
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3824607726643825,
+      "learning_rate": 1.8312865370498428e-05,
+      "loss": 0.7574,
+      "step": 2084
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.723480098849916,
+      "learning_rate": 1.8311033715739864e-05,
+      "loss": 0.7932,
+      "step": 2085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7779798532759492,
+      "learning_rate": 1.830920115894856e-05,
+      "loss": 0.7958,
+      "step": 2086
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5820203885718955,
+      "learning_rate": 1.8307367700323412e-05,
+      "loss": 0.6929,
+      "step": 2087
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6031859652237008,
+      "learning_rate": 1.8305533340063416e-05,
+      "loss": 0.7878,
+      "step": 2088
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5935160645728348,
+      "learning_rate": 1.8303698078367654e-05,
+      "loss": 0.7755,
+      "step": 2089
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5403578121776387,
+      "learning_rate": 1.8301861915435325e-05,
+      "loss": 0.8191,
+      "step": 2090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8028348646006056,
+      "learning_rate": 1.830002485146571e-05,
+      "loss": 0.8042,
+      "step": 2091
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.553546451754014,
+      "learning_rate": 1.8298186886658194e-05,
+      "loss": 0.8077,
+      "step": 2092
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5730451787934852,
+      "learning_rate": 1.8296348021212264e-05,
+      "loss": 0.8645,
+      "step": 2093
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.635196346349874,
+      "learning_rate": 1.8294508255327495e-05,
+      "loss": 0.8325,
+      "step": 2094
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7375633128416428,
+      "learning_rate": 1.8292667589203567e-05,
+      "loss": 0.7546,
+      "step": 2095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4748248166175115,
+      "learning_rate": 1.8290826023040257e-05,
+      "loss": 0.8443,
+      "step": 2096
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8397460998187334,
+      "learning_rate": 1.8288983557037432e-05,
+      "loss": 0.7881,
+      "step": 2097
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4051773976204776,
+      "learning_rate": 1.8287140191395066e-05,
+      "loss": 0.7506,
+      "step": 2098
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5543239613177329,
+      "learning_rate": 1.8285295926313234e-05,
+      "loss": 0.7805,
+      "step": 2099
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6267215636888006,
+      "learning_rate": 1.8283450761992095e-05,
+      "loss": 0.7499,
+      "step": 2100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6634829018566752,
+      "learning_rate": 1.8281604698631913e-05,
+      "loss": 0.702,
+      "step": 2101
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.458900948499807,
+      "learning_rate": 1.827975773643305e-05,
+      "loss": 0.7928,
+      "step": 2102
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7096776108046234,
+      "learning_rate": 1.8277909875595967e-05,
+      "loss": 0.7865,
+      "step": 2103
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4549477637499242,
+      "learning_rate": 1.827606111632122e-05,
+      "loss": 0.7086,
+      "step": 2104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6106913000137875,
+      "learning_rate": 1.827421145880946e-05,
+      "loss": 0.7765,
+      "step": 2105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.522056592191057,
+      "learning_rate": 1.8272360903261443e-05,
+      "loss": 0.7613,
+      "step": 2106
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.539075161245667,
+      "learning_rate": 1.8270509449878015e-05,
+      "loss": 0.724,
+      "step": 2107
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4949035601058638,
+      "learning_rate": 1.8268657098860118e-05,
+      "loss": 0.8125,
+      "step": 2108
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4720359002309913,
+      "learning_rate": 1.82668038504088e-05,
+      "loss": 0.6817,
+      "step": 2109
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.733004339510359,
+      "learning_rate": 1.826494970472521e-05,
+      "loss": 0.8364,
+      "step": 2110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5290000198905853,
+      "learning_rate": 1.8263094662010575e-05,
+      "loss": 0.7266,
+      "step": 2111
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6943982181306592,
+      "learning_rate": 1.8261238722466233e-05,
+      "loss": 0.6214,
+      "step": 2112
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7396644301300892,
+      "learning_rate": 1.825938188629362e-05,
+      "loss": 0.7582,
+      "step": 2113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.756893544882042,
+      "learning_rate": 1.8257524153694265e-05,
+      "loss": 0.7242,
+      "step": 2114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6405496596283244,
+      "learning_rate": 1.82556655248698e-05,
+      "loss": 0.8054,
+      "step": 2115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6438093570265306,
+      "learning_rate": 1.8253806000021943e-05,
+      "loss": 0.726,
+      "step": 2116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.564672755650021,
+      "learning_rate": 1.825194557935252e-05,
+      "loss": 0.6933,
+      "step": 2117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7318313364847981,
+      "learning_rate": 1.825008426306345e-05,
+      "loss": 0.7744,
+      "step": 2118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6157523650895218,
+      "learning_rate": 1.8248222051356756e-05,
+      "loss": 0.7796,
+      "step": 2119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.382492070976764,
+      "learning_rate": 1.824635894443454e-05,
+      "loss": 0.7445,
+      "step": 2120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6583871301219149,
+      "learning_rate": 1.8244494942499017e-05,
+      "loss": 0.7904,
+      "step": 2121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4231276239769954,
+      "learning_rate": 1.8242630045752504e-05,
+      "loss": 0.6558,
+      "step": 2122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.455061423838774,
+      "learning_rate": 1.8240764254397392e-05,
+      "loss": 0.7436,
+      "step": 2123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.703724941352681,
+      "learning_rate": 1.8238897568636197e-05,
+      "loss": 0.6682,
+      "step": 2124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6735807021258513,
+      "learning_rate": 1.8237029988671514e-05,
+      "loss": 0.763,
+      "step": 2125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.548686319328144,
+      "learning_rate": 1.8235161514706036e-05,
+      "loss": 0.7035,
+      "step": 2126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6797818763836223,
+      "learning_rate": 1.823329214694256e-05,
+      "loss": 0.6695,
+      "step": 2127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6404203109199544,
+      "learning_rate": 1.8231421885583972e-05,
+      "loss": 0.8632,
+      "step": 2128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5372973402500991,
+      "learning_rate": 1.822955073083327e-05,
+      "loss": 0.7308,
+      "step": 2129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6315504556007998,
+      "learning_rate": 1.822767868289353e-05,
+      "loss": 0.8769,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6043084151034726,
+      "learning_rate": 1.8225805741967934e-05,
+      "loss": 0.7148,
+      "step": 2131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5401015017426092,
+      "learning_rate": 1.822393190825976e-05,
+      "loss": 0.7348,
+      "step": 2132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4478526089039574,
+      "learning_rate": 1.8222057181972386e-05,
+      "loss": 0.7506,
+      "step": 2133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7731832083632584,
+      "learning_rate": 1.8220181563309284e-05,
+      "loss": 0.7261,
+      "step": 2134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.510479871553798,
+      "learning_rate": 1.8218305052474025e-05,
+      "loss": 0.6962,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6077436722899787,
+      "learning_rate": 1.821642764967027e-05,
+      "loss": 0.7813,
+      "step": 2136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6144631090233805,
+      "learning_rate": 1.8214549355101786e-05,
+      "loss": 0.7784,
+      "step": 2137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9735813719971183,
+      "learning_rate": 1.8212670168972428e-05,
+      "loss": 0.7713,
+      "step": 2138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6638275430587353,
+      "learning_rate": 1.8210790091486156e-05,
+      "loss": 0.7142,
+      "step": 2139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5424127528689335,
+      "learning_rate": 1.8208909122847024e-05,
+      "loss": 0.7102,
+      "step": 2140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.597958992848855,
+      "learning_rate": 1.8207027263259176e-05,
+      "loss": 0.8374,
+      "step": 2141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.565074333708335,
+      "learning_rate": 1.8205144512926866e-05,
+      "loss": 0.754,
+      "step": 2142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5897706230321191,
+      "learning_rate": 1.8203260872054432e-05,
+      "loss": 0.8083,
+      "step": 2143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4999155604503212,
+      "learning_rate": 1.8201376340846315e-05,
+      "loss": 0.6513,
+      "step": 2144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4601575616164244,
+      "learning_rate": 1.819949091950705e-05,
+      "loss": 0.6396,
+      "step": 2145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5653866484142276,
+      "learning_rate": 1.8197604608241272e-05,
+      "loss": 0.6518,
+      "step": 2146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6287065297840895,
+      "learning_rate": 1.819571740725371e-05,
+      "loss": 0.7207,
+      "step": 2147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5469126251614622,
+      "learning_rate": 1.819382931674919e-05,
+      "loss": 0.6479,
+      "step": 2148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6232718164950395,
+      "learning_rate": 1.8191940336932638e-05,
+      "loss": 0.8188,
+      "step": 2149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8874968240656473,
+      "learning_rate": 1.8190050468009065e-05,
+      "loss": 0.8426,
+      "step": 2150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6375878574874134,
+      "learning_rate": 1.8188159710183595e-05,
+      "loss": 0.831,
+      "step": 2151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6805671634638826,
+      "learning_rate": 1.8186268063661432e-05,
+      "loss": 0.7414,
+      "step": 2152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6881636311339634,
+      "learning_rate": 1.8184375528647896e-05,
+      "loss": 0.785,
+      "step": 2153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5103174375177986,
+      "learning_rate": 1.8182482105348383e-05,
+      "loss": 0.789,
+      "step": 2154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5687022363735545,
+      "learning_rate": 1.8180587793968398e-05,
+      "loss": 0.7227,
+      "step": 2155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6928307729934189,
+      "learning_rate": 1.817869259471354e-05,
+      "loss": 0.7008,
+      "step": 2156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2934833517593187,
+      "learning_rate": 1.8176796507789497e-05,
+      "loss": 0.6371,
+      "step": 2157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5592241220325302,
+      "learning_rate": 1.8174899533402066e-05,
+      "loss": 0.7327,
+      "step": 2158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8484178493317083,
+      "learning_rate": 1.8173001671757127e-05,
+      "loss": 0.8249,
+      "step": 2159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.688186929197883,
+      "learning_rate": 1.8171102923060675e-05,
+      "loss": 0.8771,
+      "step": 2160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5226220759645985,
+      "learning_rate": 1.8169203287518778e-05,
+      "loss": 0.7016,
+      "step": 2161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5748867957539676,
+      "learning_rate": 1.816730276533762e-05,
+      "loss": 0.7682,
+      "step": 2162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5209777167596974,
+      "learning_rate": 1.8165401356723467e-05,
+      "loss": 0.7089,
+      "step": 2163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5028113310786382,
+      "learning_rate": 1.816349906188269e-05,
+      "loss": 0.6731,
+      "step": 2164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.572176249001475,
+      "learning_rate": 1.8161595881021756e-05,
+      "loss": 0.8238,
+      "step": 2165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5882665731586676,
+      "learning_rate": 1.8159691814347217e-05,
+      "loss": 0.8786,
+      "step": 2166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5656827837258642,
+      "learning_rate": 1.8157786862065735e-05,
+      "loss": 0.7759,
+      "step": 2167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4514364376988993,
+      "learning_rate": 1.8155881024384064e-05,
+      "loss": 0.835,
+      "step": 2168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.588171207975576,
+      "learning_rate": 1.8153974301509053e-05,
+      "loss": 0.7176,
+      "step": 2169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7084577912367054,
+      "learning_rate": 1.815206669364764e-05,
+      "loss": 0.8163,
+      "step": 2170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5668159027685062,
+      "learning_rate": 1.8150158201006878e-05,
+      "loss": 0.8995,
+      "step": 2171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8305587750700543,
+      "learning_rate": 1.8148248823793895e-05,
+      "loss": 0.7976,
+      "step": 2172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5635711829004972,
+      "learning_rate": 1.8146338562215927e-05,
+      "loss": 0.8354,
+      "step": 2173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.569805517173065,
+      "learning_rate": 1.81444274164803e-05,
+      "loss": 0.8663,
+      "step": 2174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.583628255647632,
+      "learning_rate": 1.8142515386794443e-05,
+      "loss": 0.7238,
+      "step": 2175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5298895392029468,
+      "learning_rate": 1.814060247336588e-05,
+      "loss": 0.7605,
+      "step": 2176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6327355633469391,
+      "learning_rate": 1.8138688676402212e-05,
+      "loss": 0.7277,
+      "step": 2177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4512353475385125,
+      "learning_rate": 1.8136773996111175e-05,
+      "loss": 0.6717,
+      "step": 2178
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7808283545662742,
+      "learning_rate": 1.813485843270056e-05,
+      "loss": 0.8072,
+      "step": 2179
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6894019878441329,
+      "learning_rate": 1.8132941986378276e-05,
+      "loss": 0.6918,
+      "step": 2180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2985355768151867,
+      "learning_rate": 1.8131024657352328e-05,
+      "loss": 0.7033,
+      "step": 2181
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3873095945563672,
+      "learning_rate": 1.8129106445830807e-05,
+      "loss": 0.8286,
+      "step": 2182
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7586578364971428,
+      "learning_rate": 1.8127187352021908e-05,
+      "loss": 0.8945,
+      "step": 2183
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6220564789546164,
+      "learning_rate": 1.8125267376133912e-05,
+      "loss": 0.81,
+      "step": 2184
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6101531248573264,
+      "learning_rate": 1.812334651837521e-05,
+      "loss": 0.7452,
+      "step": 2185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5883461039783242,
+      "learning_rate": 1.8121424778954278e-05,
+      "loss": 0.8675,
+      "step": 2186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4514916706150318,
+      "learning_rate": 1.8119502158079693e-05,
+      "loss": 0.7217,
+      "step": 2187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6195550361174957,
+      "learning_rate": 1.8117578655960123e-05,
+      "loss": 0.9525,
+      "step": 2188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5460650451901794,
+      "learning_rate": 1.8115654272804333e-05,
+      "loss": 0.6632,
+      "step": 2189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5250281746069403,
+      "learning_rate": 1.811372900882119e-05,
+      "loss": 0.6274,
+      "step": 2190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5412167993310468,
+      "learning_rate": 1.8111802864219643e-05,
+      "loss": 0.7328,
+      "step": 2191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.637075226598729,
+      "learning_rate": 1.8109875839208754e-05,
+      "loss": 0.7426,
+      "step": 2192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6717734330324852,
+      "learning_rate": 1.8107947933997663e-05,
+      "loss": 0.8137,
+      "step": 2193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.530759629670791,
+      "learning_rate": 1.810601914879562e-05,
+      "loss": 0.7175,
+      "step": 2194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.526483843980927,
+      "learning_rate": 1.8104089483811963e-05,
+      "loss": 0.7751,
+      "step": 2195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5835310478088402,
+      "learning_rate": 1.8102158939256122e-05,
+      "loss": 0.7437,
+      "step": 2196
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6780789997117656,
+      "learning_rate": 1.8100227515337634e-05,
+      "loss": 0.7519,
+      "step": 2197
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6293693684949369,
+      "learning_rate": 1.8098295212266123e-05,
+      "loss": 0.7471,
+      "step": 2198
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5963838193341875,
+      "learning_rate": 1.8096362030251312e-05,
+      "loss": 0.8278,
+      "step": 2199
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.641469573490994,
+      "learning_rate": 1.8094427969503013e-05,
+      "loss": 0.772,
+      "step": 2200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.634557774448412,
+      "learning_rate": 1.8092493030231142e-05,
+      "loss": 0.8353,
+      "step": 2201
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.565183800023239,
+      "learning_rate": 1.8090557212645702e-05,
+      "loss": 0.7399,
+      "step": 2202
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7881324029802745,
+      "learning_rate": 1.8088620516956804e-05,
+      "loss": 0.8626,
+      "step": 2203
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4944277512903883,
+      "learning_rate": 1.808668294337464e-05,
+      "loss": 0.7774,
+      "step": 2204
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.527912053127469,
+      "learning_rate": 1.8084744492109497e-05,
+      "loss": 0.7545,
+      "step": 2205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.592145236674976,
+      "learning_rate": 1.8082805163371777e-05,
+      "loss": 0.7721,
+      "step": 2206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6812339574309858,
+      "learning_rate": 1.8080864957371958e-05,
+      "loss": 0.7945,
+      "step": 2207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6211980361220193,
+      "learning_rate": 1.8078923874320615e-05,
+      "loss": 0.7518,
+      "step": 2208
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.614036868443832,
+      "learning_rate": 1.8076981914428426e-05,
+      "loss": 0.868,
+      "step": 2209
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6792941191195823,
+      "learning_rate": 1.8075039077906163e-05,
+      "loss": 0.8067,
+      "step": 2210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5699987593174205,
+      "learning_rate": 1.8073095364964682e-05,
+      "loss": 0.7725,
+      "step": 2211
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5710191725635028,
+      "learning_rate": 1.8071150775814956e-05,
+      "loss": 0.8535,
+      "step": 2212
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6947804590122968,
+      "learning_rate": 1.8069205310668026e-05,
+      "loss": 0.7741,
+      "step": 2213
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4980756469697138,
+      "learning_rate": 1.8067258969735052e-05,
+      "loss": 0.727,
+      "step": 2214
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.452475006388767,
+      "learning_rate": 1.8065311753227272e-05,
+      "loss": 0.6818,
+      "step": 2215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5605286651538655,
+      "learning_rate": 1.8063363661356027e-05,
+      "loss": 0.7793,
+      "step": 2216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6668571113325545,
+      "learning_rate": 1.8061414694332756e-05,
+      "loss": 0.6445,
+      "step": 2217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5436010718479045,
+      "learning_rate": 1.8059464852368987e-05,
+      "loss": 0.8266,
+      "step": 2218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5529862587308387,
+      "learning_rate": 1.8057514135676344e-05,
+      "loss": 0.6925,
+      "step": 2219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.673046218654071,
+      "learning_rate": 1.8055562544466544e-05,
+      "loss": 0.7221,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6071811351863372,
+      "learning_rate": 1.8053610078951406e-05,
+      "loss": 0.7382,
+      "step": 2221
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.697397817656138,
+      "learning_rate": 1.805165673934284e-05,
+      "loss": 0.8095,
+      "step": 2222
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.494326366333854,
+      "learning_rate": 1.804970252585285e-05,
+      "loss": 0.7678,
+      "step": 2223
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.460052000381731,
+      "learning_rate": 1.804774743869353e-05,
+      "loss": 0.6809,
+      "step": 2224
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6463863753897041,
+      "learning_rate": 1.804579147807708e-05,
+      "loss": 0.9014,
+      "step": 2225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4916566202877668,
+      "learning_rate": 1.8043834644215788e-05,
+      "loss": 0.7232,
+      "step": 2226
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4792196205038426,
+      "learning_rate": 1.8041876937322037e-05,
+      "loss": 0.7147,
+      "step": 2227
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6681977826488052,
+      "learning_rate": 1.8039918357608304e-05,
+      "loss": 0.7517,
+      "step": 2228
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6038026551810642,
+      "learning_rate": 1.803795890528716e-05,
+      "loss": 0.7602,
+      "step": 2229
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6556802077399546,
+      "learning_rate": 1.8035998580571282e-05,
+      "loss": 0.6506,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.738570205110402,
+      "learning_rate": 1.8034037383673428e-05,
+      "loss": 0.7763,
+      "step": 2231
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5473612742580263,
+      "learning_rate": 1.803207531480645e-05,
+      "loss": 0.6568,
+      "step": 2232
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6939965516749653,
+      "learning_rate": 1.8030112374183308e-05,
+      "loss": 0.7322,
+      "step": 2233
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.528914442618967,
+      "learning_rate": 1.802814856201704e-05,
+      "loss": 0.7585,
+      "step": 2234
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4757461250675585,
+      "learning_rate": 1.8026183878520794e-05,
+      "loss": 0.7902,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5769224453410504,
+      "learning_rate": 1.8024218323907807e-05,
+      "loss": 0.6148,
+      "step": 2236
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4609111499910967,
+      "learning_rate": 1.8022251898391402e-05,
+      "loss": 0.6959,
+      "step": 2237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.434729618799995,
+      "learning_rate": 1.8020284602185006e-05,
+      "loss": 0.7305,
+      "step": 2238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4238232846508847,
+      "learning_rate": 1.8018316435502143e-05,
+      "loss": 0.6863,
+      "step": 2239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5818552847046934,
+      "learning_rate": 1.8016347398556423e-05,
+      "loss": 0.7345,
+      "step": 2240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7039010405467567,
+      "learning_rate": 1.8014377491561553e-05,
+      "loss": 0.7224,
+      "step": 2241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6340705954608221,
+      "learning_rate": 1.8012406714731338e-05,
+      "loss": 0.8247,
+      "step": 2242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5826030085527707,
+      "learning_rate": 1.8010435068279673e-05,
+      "loss": 0.8059,
+      "step": 2243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5847457915547827,
+      "learning_rate": 1.800846255242055e-05,
+      "loss": 0.7552,
+      "step": 2244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7361257789569242,
+      "learning_rate": 1.800648916736806e-05,
+      "loss": 0.7518,
+      "step": 2245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7586600117640085,
+      "learning_rate": 1.8004514913336374e-05,
+      "loss": 0.9412,
+      "step": 2246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6696786408496551,
+      "learning_rate": 1.8002539790539772e-05,
+      "loss": 0.8753,
+      "step": 2247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.559155976771375,
+      "learning_rate": 1.8000563799192622e-05,
+      "loss": 0.7045,
+      "step": 2248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5477534172888618,
+      "learning_rate": 1.7998586939509385e-05,
+      "loss": 0.7251,
+      "step": 2249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7697877319548974,
+      "learning_rate": 1.799660921170462e-05,
+      "loss": 0.8147,
+      "step": 2250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.629767247929377,
+      "learning_rate": 1.799463061599298e-05,
+      "loss": 0.9108,
+      "step": 2251
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.596274521142993,
+      "learning_rate": 1.7992651152589205e-05,
+      "loss": 0.7334,
+      "step": 2252
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6516715835230245,
+      "learning_rate": 1.799067082170814e-05,
+      "loss": 0.7745,
+      "step": 2253
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8505464121141373,
+      "learning_rate": 1.798868962356472e-05,
+      "loss": 0.7958,
+      "step": 2254
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5896082359240096,
+      "learning_rate": 1.798670755837397e-05,
+      "loss": 0.7953,
+      "step": 2255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.574272206265239,
+      "learning_rate": 1.7984724626351008e-05,
+      "loss": 0.8106,
+      "step": 2256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.558125079934259,
+      "learning_rate": 1.7982740827711058e-05,
+      "loss": 0.6858,
+      "step": 2257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5459963520716786,
+      "learning_rate": 1.7980756162669427e-05,
+      "loss": 0.7438,
+      "step": 2258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5282081192747206,
+      "learning_rate": 1.797877063144152e-05,
+      "loss": 0.7825,
+      "step": 2259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.523370901061267,
+      "learning_rate": 1.7976784234242838e-05,
+      "loss": 0.6955,
+      "step": 2260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5460023446118578,
+      "learning_rate": 1.7974796971288966e-05,
+      "loss": 0.7258,
+      "step": 2261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3836104999840484,
+      "learning_rate": 1.7972808842795594e-05,
+      "loss": 0.8183,
+      "step": 2262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6467488935328332,
+      "learning_rate": 1.7970819848978503e-05,
+      "loss": 0.7974,
+      "step": 2263
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.513920497223266,
+      "learning_rate": 1.7968829990053572e-05,
+      "loss": 0.7312,
+      "step": 2264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5416125278602086,
+      "learning_rate": 1.796683926623676e-05,
+      "loss": 0.8179,
+      "step": 2265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6108320417733,
+      "learning_rate": 1.7964847677744136e-05,
+      "loss": 0.719,
+      "step": 2266
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5754808386958055,
+      "learning_rate": 1.7962855224791852e-05,
+      "loss": 0.7511,
+      "step": 2267
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4482906664508932,
+      "learning_rate": 1.7960861907596158e-05,
+      "loss": 0.7618,
+      "step": 2268
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5903384355437244,
+      "learning_rate": 1.7958867726373393e-05,
+      "loss": 0.7502,
+      "step": 2269
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3129301228336305,
+      "learning_rate": 1.7956872681340005e-05,
+      "loss": 0.7442,
+      "step": 2270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.60673569630705,
+      "learning_rate": 1.795487677271252e-05,
+      "loss": 0.7776,
+      "step": 2271
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.655070352685222,
+      "learning_rate": 1.7952880000707558e-05,
+      "loss": 0.6787,
+      "step": 2272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5340526168662216,
+      "learning_rate": 1.795088236554184e-05,
+      "loss": 0.7754,
+      "step": 2273
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.525057887506206,
+      "learning_rate": 1.7948883867432184e-05,
+      "loss": 0.8401,
+      "step": 2274
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6525453707981543,
+      "learning_rate": 1.7946884506595487e-05,
+      "loss": 0.7541,
+      "step": 2275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6554115210034996,
+      "learning_rate": 1.794488428324875e-05,
+      "loss": 0.7034,
+      "step": 2276
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4414860762539807,
+      "learning_rate": 1.7942883197609074e-05,
+      "loss": 0.6702,
+      "step": 2277
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.594305792624561,
+      "learning_rate": 1.7940881249893636e-05,
+      "loss": 0.7537,
+      "step": 2278
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6752368175096224,
+      "learning_rate": 1.7938878440319722e-05,
+      "loss": 0.8536,
+      "step": 2279
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3880591018903257,
+      "learning_rate": 1.7936874769104703e-05,
+      "loss": 0.5761,
+      "step": 2280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6830662990808718,
+      "learning_rate": 1.7934870236466047e-05,
+      "loss": 0.7947,
+      "step": 2281
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5170592285262723,
+      "learning_rate": 1.7932864842621312e-05,
+      "loss": 0.7356,
+      "step": 2282
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.590286210609126,
+      "learning_rate": 1.7930858587788156e-05,
+      "loss": 0.7456,
+      "step": 2283
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5899046321067252,
+      "learning_rate": 1.7928851472184323e-05,
+      "loss": 0.7471,
+      "step": 2284
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.830304910240319,
+      "learning_rate": 1.792684349602766e-05,
+      "loss": 0.8573,
+      "step": 2285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5360155713568722,
+      "learning_rate": 1.7924834659536093e-05,
+      "loss": 0.899,
+      "step": 2286
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.503268262966366,
+      "learning_rate": 1.7922824962927657e-05,
+      "loss": 0.808,
+      "step": 2287
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6637303000309078,
+      "learning_rate": 1.792081440642047e-05,
+      "loss": 0.8694,
+      "step": 2288
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4930490566181709,
+      "learning_rate": 1.7918802990232748e-05,
+      "loss": 0.7238,
+      "step": 2289
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6580291703719623,
+      "learning_rate": 1.7916790714582795e-05,
+      "loss": 0.7738,
+      "step": 2290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5967606389216442,
+      "learning_rate": 1.7914777579689017e-05,
+      "loss": 0.7241,
+      "step": 2291
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6330398542636713,
+      "learning_rate": 1.7912763585769905e-05,
+      "loss": 0.8487,
+      "step": 2292
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4311002448815942,
+      "learning_rate": 1.791074873304405e-05,
+      "loss": 0.7219,
+      "step": 2293
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.738071137245182,
+      "learning_rate": 1.790873302173013e-05,
+      "loss": 0.7444,
+      "step": 2294
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5850740120003257,
+      "learning_rate": 1.7906716452046915e-05,
+      "loss": 0.7367,
+      "step": 2295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5592420662588853,
+      "learning_rate": 1.790469902421328e-05,
+      "loss": 0.7313,
+      "step": 2296
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4806536703150404,
+      "learning_rate": 1.7902680738448178e-05,
+      "loss": 0.8097,
+      "step": 2297
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.489296277259454,
+      "learning_rate": 1.790066159497067e-05,
+      "loss": 0.7146,
+      "step": 2298
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.558413346803114,
+      "learning_rate": 1.7898641593999893e-05,
+      "loss": 0.7599,
+      "step": 2299
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4735137697579415,
+      "learning_rate": 1.78966207357551e-05,
+      "loss": 0.8103,
+      "step": 2300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4504295184011107,
+      "learning_rate": 1.789459902045561e-05,
+      "loss": 0.6457,
+      "step": 2301
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4983891556182687,
+      "learning_rate": 1.7892576448320854e-05,
+      "loss": 0.7301,
+      "step": 2302
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.635941334049674,
+      "learning_rate": 1.7890553019570353e-05,
+      "loss": 0.7768,
+      "step": 2303
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3864888565813458,
+      "learning_rate": 1.7888528734423715e-05,
+      "loss": 0.6055,
+      "step": 2304
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4703550242509769,
+      "learning_rate": 1.7886503593100646e-05,
+      "loss": 0.7608,
+      "step": 2305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6246563516863506,
+      "learning_rate": 1.788447759582094e-05,
+      "loss": 0.7621,
+      "step": 2306
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.577237508115445,
+      "learning_rate": 1.7882450742804493e-05,
+      "loss": 0.7709,
+      "step": 2307
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.690658860404219,
+      "learning_rate": 1.7880423034271287e-05,
+      "loss": 0.7846,
+      "step": 2308
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6931476627904156,
+      "learning_rate": 1.7878394470441396e-05,
+      "loss": 0.6558,
+      "step": 2309
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6204463011198105,
+      "learning_rate": 1.7876365051534987e-05,
+      "loss": 0.8227,
+      "step": 2310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.522142949127196,
+      "learning_rate": 1.7874334777772326e-05,
+      "loss": 0.7449,
+      "step": 2311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6461329249430225,
+      "learning_rate": 1.7872303649373767e-05,
+      "loss": 0.7682,
+      "step": 2312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5824110708824122,
+      "learning_rate": 1.7870271666559755e-05,
+      "loss": 0.7866,
+      "step": 2313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6400074273266572,
+      "learning_rate": 1.786823882955083e-05,
+      "loss": 0.8164,
+      "step": 2314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.502158375278795,
+      "learning_rate": 1.7866205138567626e-05,
+      "loss": 0.7833,
+      "step": 2315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4295665349426534,
+      "learning_rate": 1.7864170593830868e-05,
+      "loss": 0.7662,
+      "step": 2316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.516678854068183,
+      "learning_rate": 1.7862135195561373e-05,
+      "loss": 0.6889,
+      "step": 2317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.546924044213888,
+      "learning_rate": 1.7860098943980056e-05,
+      "loss": 0.6845,
+      "step": 2318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5686659936839316,
+      "learning_rate": 1.7858061839307913e-05,
+      "loss": 0.7132,
+      "step": 2319
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5550197845969858,
+      "learning_rate": 1.7856023881766048e-05,
+      "loss": 0.7091,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.529551763484397,
+      "learning_rate": 1.785398507157564e-05,
+      "loss": 0.7918,
+      "step": 2321
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6061716867769542,
+      "learning_rate": 1.7851945408957977e-05,
+      "loss": 0.788,
+      "step": 2322
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.556076449285782,
+      "learning_rate": 1.784990489413443e-05,
+      "loss": 0.8446,
+      "step": 2323
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6842789691197135,
+      "learning_rate": 1.7847863527326468e-05,
+      "loss": 0.7131,
+      "step": 2324
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5732131853316018,
+      "learning_rate": 1.7845821308755644e-05,
+      "loss": 0.7275,
+      "step": 2325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6710459182156756,
+      "learning_rate": 1.7843778238643615e-05,
+      "loss": 0.7791,
+      "step": 2326
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.767202341467178,
+      "learning_rate": 1.784173431721212e-05,
+      "loss": 0.801,
+      "step": 2327
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5185138048879634,
+      "learning_rate": 1.7839689544682993e-05,
+      "loss": 0.8229,
+      "step": 2328
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6198534151063857,
+      "learning_rate": 1.7837643921278168e-05,
+      "loss": 0.763,
+      "step": 2329
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3949632519487039,
+      "learning_rate": 1.783559744721966e-05,
+      "loss": 0.7574,
+      "step": 2330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6805417262534503,
+      "learning_rate": 1.7833550122729587e-05,
+      "loss": 0.728,
+      "step": 2331
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5531944312041643,
+      "learning_rate": 1.783150194803015e-05,
+      "loss": 0.7335,
+      "step": 2332
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7159353418847874,
+      "learning_rate": 1.7829452923343648e-05,
+      "loss": 0.632,
+      "step": 2333
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.491814756386104,
+      "learning_rate": 1.782740304889247e-05,
+      "loss": 0.7295,
+      "step": 2334
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6304216530256852,
+      "learning_rate": 1.7825352324899093e-05,
+      "loss": 0.7432,
+      "step": 2335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5059743877915686,
+      "learning_rate": 1.7823300751586103e-05,
+      "loss": 0.8248,
+      "step": 2336
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5372983039722445,
+      "learning_rate": 1.7821248329176156e-05,
+      "loss": 0.7279,
+      "step": 2337
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.634906198136157,
+      "learning_rate": 1.7819195057892015e-05,
+      "loss": 0.8099,
+      "step": 2338
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.395598233074499,
+      "learning_rate": 1.781714093795653e-05,
+      "loss": 0.7467,
+      "step": 2339
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.387677729120744,
+      "learning_rate": 1.781508596959264e-05,
+      "loss": 0.6657,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.665462662562675,
+      "learning_rate": 1.7813030153023382e-05,
+      "loss": 0.7382,
+      "step": 2341
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4919226889245358,
+      "learning_rate": 1.7810973488471882e-05,
+      "loss": 0.8811,
+      "step": 2342
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.570893636073215,
+      "learning_rate": 1.7808915976161364e-05,
+      "loss": 0.6923,
+      "step": 2343
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.574635587846645,
+      "learning_rate": 1.780685761631513e-05,
+      "loss": 0.7827,
+      "step": 2344
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6669870967013605,
+      "learning_rate": 1.7804798409156592e-05,
+      "loss": 0.8147,
+      "step": 2345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7019734330167111,
+      "learning_rate": 1.7802738354909236e-05,
+      "loss": 0.8692,
+      "step": 2346
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5679652330468707,
+      "learning_rate": 1.7800677453796656e-05,
+      "loss": 0.7307,
+      "step": 2347
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5464299583191818,
+      "learning_rate": 1.7798615706042525e-05,
+      "loss": 0.6894,
+      "step": 2348
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5136573641374516,
+      "learning_rate": 1.7796553111870616e-05,
+      "loss": 0.7219,
+      "step": 2349
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4193934300294413,
+      "learning_rate": 1.7794489671504793e-05,
+      "loss": 0.7493,
+      "step": 2350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4403542637862665,
+      "learning_rate": 1.7792425385169006e-05,
+      "loss": 0.7527,
+      "step": 2351
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7784718114779616,
+      "learning_rate": 1.7790360253087304e-05,
+      "loss": 0.6761,
+      "step": 2352
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5741525022384846,
+      "learning_rate": 1.7788294275483826e-05,
+      "loss": 0.7737,
+      "step": 2353
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5300340210520655,
+      "learning_rate": 1.77862274525828e-05,
+      "loss": 0.7701,
+      "step": 2354
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.668962554632878,
+      "learning_rate": 1.7784159784608543e-05,
+      "loss": 0.7207,
+      "step": 2355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4913381646840573,
+      "learning_rate": 1.7782091271785475e-05,
+      "loss": 0.6585,
+      "step": 2356
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6479277771557135,
+      "learning_rate": 1.7780021914338097e-05,
+      "loss": 0.7769,
+      "step": 2357
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5844220072583104,
+      "learning_rate": 1.7777951712491007e-05,
+      "loss": 0.7385,
+      "step": 2358
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5974427028035825,
+      "learning_rate": 1.7775880666468893e-05,
+      "loss": 0.7192,
+      "step": 2359
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5697418827414715,
+      "learning_rate": 1.777380877649653e-05,
+      "loss": 0.7989,
+      "step": 2360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6440250115104977,
+      "learning_rate": 1.7771736042798796e-05,
+      "loss": 0.8351,
+      "step": 2361
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.641336175166559,
+      "learning_rate": 1.7769662465600648e-05,
+      "loss": 0.9197,
+      "step": 2362
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4982313511845302,
+      "learning_rate": 1.776758804512715e-05,
+      "loss": 0.783,
+      "step": 2363
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6681412333189058,
+      "learning_rate": 1.7765512781603434e-05,
+      "loss": 0.8278,
+      "step": 2364
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5493644062862706,
+      "learning_rate": 1.776343667525475e-05,
+      "loss": 0.6787,
+      "step": 2365
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6440060561290857,
+      "learning_rate": 1.7761359726306416e-05,
+      "loss": 0.682,
+      "step": 2366
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6511456873028476,
+      "learning_rate": 1.775928193498386e-05,
+      "loss": 0.8085,
+      "step": 2367
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6045551984799362,
+      "learning_rate": 1.775720330151259e-05,
+      "loss": 0.7526,
+      "step": 2368
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5439942621819842,
+      "learning_rate": 1.7755123826118215e-05,
+      "loss": 0.8021,
+      "step": 2369
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4369959230768674,
+      "learning_rate": 1.7753043509026423e-05,
+      "loss": 0.7637,
+      "step": 2370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6486282948075461,
+      "learning_rate": 1.7750962350463005e-05,
+      "loss": 0.8355,
+      "step": 2371
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6284762608456098,
+      "learning_rate": 1.774888035065383e-05,
+      "loss": 0.8257,
+      "step": 2372
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5193088564965311,
+      "learning_rate": 1.7746797509824875e-05,
+      "loss": 0.7028,
+      "step": 2373
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4951644348597182,
+      "learning_rate": 1.7744713828202197e-05,
+      "loss": 0.7792,
+      "step": 2374
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6373593823681687,
+      "learning_rate": 1.7742629306011944e-05,
+      "loss": 0.7483,
+      "step": 2375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5588424733699444,
+      "learning_rate": 1.7740543943480366e-05,
+      "loss": 0.7492,
+      "step": 2376
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5793556324853542,
+      "learning_rate": 1.7738457740833785e-05,
+      "loss": 0.7173,
+      "step": 2377
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5671484025185511,
+      "learning_rate": 1.7736370698298637e-05,
+      "loss": 0.7549,
+      "step": 2378
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6090619943711126,
+      "learning_rate": 1.773428281610143e-05,
+      "loss": 0.8098,
+      "step": 2379
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5578260997102495,
+      "learning_rate": 1.7732194094468774e-05,
+      "loss": 0.7443,
+      "step": 2380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4643681684258554,
+      "learning_rate": 1.773010453362737e-05,
+      "loss": 0.6952,
+      "step": 2381
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6443043569457474,
+      "learning_rate": 1.7728014133804004e-05,
+      "loss": 0.7982,
+      "step": 2382
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7237283774975798,
+      "learning_rate": 1.7725922895225554e-05,
+      "loss": 0.7161,
+      "step": 2383
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5062277487135245,
+      "learning_rate": 1.7723830818118997e-05,
+      "loss": 0.8251,
+      "step": 2384
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5078731457102554,
+      "learning_rate": 1.772173790271139e-05,
+      "loss": 0.6404,
+      "step": 2385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5321573853321941,
+      "learning_rate": 1.7719644149229888e-05,
+      "loss": 0.7963,
+      "step": 2386
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4672533371132384,
+      "learning_rate": 1.7717549557901735e-05,
+      "loss": 0.6917,
+      "step": 2387
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6271594580379487,
+      "learning_rate": 1.7715454128954265e-05,
+      "loss": 0.8336,
+      "step": 2388
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5602583919191477,
+      "learning_rate": 1.771335786261491e-05,
+      "loss": 0.7374,
+      "step": 2389
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5818657874167723,
+      "learning_rate": 1.7711260759111176e-05,
+      "loss": 0.828,
+      "step": 2390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.558116204908089,
+      "learning_rate": 1.7709162818670682e-05,
+      "loss": 0.6835,
+      "step": 2391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7432452033050492,
+      "learning_rate": 1.770706404152112e-05,
+      "loss": 0.8202,
+      "step": 2392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5211509608944633,
+      "learning_rate": 1.7704964427890283e-05,
+      "loss": 0.7055,
+      "step": 2393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5952756302093576,
+      "learning_rate": 1.770286397800605e-05,
+      "loss": 0.6794,
+      "step": 2394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5888107698469187,
+      "learning_rate": 1.770076269209639e-05,
+      "loss": 0.7322,
+      "step": 2395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5430566750829608,
+      "learning_rate": 1.7698660570389362e-05,
+      "loss": 0.79,
+      "step": 2396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.657367611936941,
+      "learning_rate": 1.769655761311313e-05,
+      "loss": 0.7225,
+      "step": 2397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5974172027320854,
+      "learning_rate": 1.7694453820495926e-05,
+      "loss": 0.8598,
+      "step": 2398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5034867091666928,
+      "learning_rate": 1.769234919276609e-05,
+      "loss": 0.7784,
+      "step": 2399
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5627926366672371,
+      "learning_rate": 1.7690243730152043e-05,
+      "loss": 0.8084,
+      "step": 2400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5433255789976013,
+      "learning_rate": 1.76881374328823e-05,
+      "loss": 0.7259,
+      "step": 2401
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4802198654414553,
+      "learning_rate": 1.7686030301185467e-05,
+      "loss": 0.7845,
+      "step": 2402
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5150080798011463,
+      "learning_rate": 1.768392233529024e-05,
+      "loss": 0.8206,
+      "step": 2403
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.443068138380811,
+      "learning_rate": 1.768181353542541e-05,
+      "loss": 0.7574,
+      "step": 2404
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6784448140480828,
+      "learning_rate": 1.7679703901819846e-05,
+      "loss": 0.7368,
+      "step": 2405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6090442224123616,
+      "learning_rate": 1.7677593434702525e-05,
+      "loss": 0.7393,
+      "step": 2406
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5542583164937114,
+      "learning_rate": 1.7675482134302503e-05,
+      "loss": 0.8111,
+      "step": 2407
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.471717033418376,
+      "learning_rate": 1.767337000084892e-05,
+      "loss": 0.8439,
+      "step": 2408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7737988262489541,
+      "learning_rate": 1.7671257034571027e-05,
+      "loss": 0.726,
+      "step": 2409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5441726699663372,
+      "learning_rate": 1.7669143235698147e-05,
+      "loss": 0.7305,
+      "step": 2410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6664292775600074,
+      "learning_rate": 1.7667028604459698e-05,
+      "loss": 0.8451,
+      "step": 2411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6825398377815297,
+      "learning_rate": 1.7664913141085192e-05,
+      "loss": 0.8154,
+      "step": 2412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6625355468496055,
+      "learning_rate": 1.7662796845804237e-05,
+      "loss": 0.7677,
+      "step": 2413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6267473749888721,
+      "learning_rate": 1.766067971884651e-05,
+      "loss": 0.7411,
+      "step": 2414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5825823341562621,
+      "learning_rate": 1.7658561760441803e-05,
+      "loss": 0.7802,
+      "step": 2415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6240718740048647,
+      "learning_rate": 1.765644297081998e-05,
+      "loss": 0.7313,
+      "step": 2416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6696875876375654,
+      "learning_rate": 1.7654323350211008e-05,
+      "loss": 0.7813,
+      "step": 2417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5501823094293559,
+      "learning_rate": 1.7652202898844935e-05,
+      "loss": 0.7431,
+      "step": 2418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.50758234605498,
+      "learning_rate": 1.765008161695191e-05,
+      "loss": 0.6745,
+      "step": 2419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4204801932106466,
+      "learning_rate": 1.7647959504762155e-05,
+      "loss": 0.6802,
+      "step": 2420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5743749498683806,
+      "learning_rate": 1.7645836562506e-05,
+      "loss": 0.7562,
+      "step": 2421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3687332132205343,
+      "learning_rate": 1.7643712790413848e-05,
+      "loss": 0.5944,
+      "step": 2422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7182075046051772,
+      "learning_rate": 1.764158818871621e-05,
+      "loss": 0.7734,
+      "step": 2423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5626744617095825,
+      "learning_rate": 1.7639462757643672e-05,
+      "loss": 0.75,
+      "step": 2424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5197050336323787,
+      "learning_rate": 1.7637336497426925e-05,
+      "loss": 0.7943,
+      "step": 2425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3619574222041446,
+      "learning_rate": 1.7635209408296733e-05,
+      "loss": 0.6312,
+      "step": 2426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6305415758528246,
+      "learning_rate": 1.763308149048396e-05,
+      "loss": 0.888,
+      "step": 2427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5546492520862543,
+      "learning_rate": 1.763095274421956e-05,
+      "loss": 0.8019,
+      "step": 2428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.555215142648058,
+      "learning_rate": 1.7628823169734573e-05,
+      "loss": 0.7817,
+      "step": 2429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.511727865036425,
+      "learning_rate": 1.7626692767260136e-05,
+      "loss": 0.7247,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.804382355917543,
+      "learning_rate": 1.7624561537027465e-05,
+      "loss": 0.8425,
+      "step": 2431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6648886739946853,
+      "learning_rate": 1.7622429479267876e-05,
+      "loss": 0.8356,
+      "step": 2432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7135054130486669,
+      "learning_rate": 1.7620296594212766e-05,
+      "loss": 0.7348,
+      "step": 2433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4794522885096486,
+      "learning_rate": 1.761816288209363e-05,
+      "loss": 0.7983,
+      "step": 2434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7887338748656285,
+      "learning_rate": 1.7616028343142048e-05,
+      "loss": 0.8899,
+      "step": 2435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.630679636137063,
+      "learning_rate": 1.7613892977589692e-05,
+      "loss": 0.8243,
+      "step": 2436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.625913634437561,
+      "learning_rate": 1.7611756785668313e-05,
+      "loss": 0.823,
+      "step": 2437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4996521582223123,
+      "learning_rate": 1.7609619767609782e-05,
+      "loss": 0.7152,
+      "step": 2438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6813915627504108,
+      "learning_rate": 1.760748192364602e-05,
+      "loss": 0.8351,
+      "step": 2439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.563039082175078,
+      "learning_rate": 1.7605343254009063e-05,
+      "loss": 0.82,
+      "step": 2440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5668253334856597,
+      "learning_rate": 1.7603203758931027e-05,
+      "loss": 0.6734,
+      "step": 2441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.64218612244895,
+      "learning_rate": 1.760106343864413e-05,
+      "loss": 0.7117,
+      "step": 2442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4172340689668361,
+      "learning_rate": 1.759892229338066e-05,
+      "loss": 0.7138,
+      "step": 2443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5161626935295773,
+      "learning_rate": 1.759678032337301e-05,
+      "loss": 0.7911,
+      "step": 2444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6555376454889454,
+      "learning_rate": 1.7594637528853654e-05,
+      "loss": 0.7448,
+      "step": 2445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6448450279141316,
+      "learning_rate": 1.7592493910055165e-05,
+      "loss": 0.8029,
+      "step": 2446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6811458558176975,
+      "learning_rate": 1.759034946721019e-05,
+      "loss": 0.8106,
+      "step": 2447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4864518399180122,
+      "learning_rate": 1.7588204200551486e-05,
+      "loss": 0.8088,
+      "step": 2448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.685284783331612,
+      "learning_rate": 1.7586058110311882e-05,
+      "loss": 0.7551,
+      "step": 2449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7535255994342807,
+      "learning_rate": 1.75839111967243e-05,
+      "loss": 0.7497,
+      "step": 2450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.519266239438577,
+      "learning_rate": 1.7581763460021758e-05,
+      "loss": 0.7114,
+      "step": 2451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5205905936797746,
+      "learning_rate": 1.757961490043736e-05,
+      "loss": 0.7277,
+      "step": 2452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4832968313005075,
+      "learning_rate": 1.757746551820429e-05,
+      "loss": 0.6313,
+      "step": 2453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5906751619604762,
+      "learning_rate": 1.7575315313555844e-05,
+      "loss": 0.7671,
+      "step": 2454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5306138592197545,
+      "learning_rate": 1.757316428672538e-05,
+      "loss": 0.8233,
+      "step": 2455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5420758259052774,
+      "learning_rate": 1.7571012437946365e-05,
+      "loss": 0.8207,
+      "step": 2456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4863632131027524,
+      "learning_rate": 1.7568859767452347e-05,
+      "loss": 0.6853,
+      "step": 2457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5257917166953734,
+      "learning_rate": 1.7566706275476967e-05,
+      "loss": 0.6817,
+      "step": 2458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4743887619029097,
+      "learning_rate": 1.7564551962253952e-05,
+      "loss": 0.7811,
+      "step": 2459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5472620487297755,
+      "learning_rate": 1.756239682801711e-05,
+      "loss": 0.7322,
+      "step": 2460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4691655852007381,
+      "learning_rate": 1.7560240873000363e-05,
+      "loss": 0.7562,
+      "step": 2461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4818530953336322,
+      "learning_rate": 1.7558084097437697e-05,
+      "loss": 0.6782,
+      "step": 2462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5238066160364558,
+      "learning_rate": 1.7555926501563198e-05,
+      "loss": 0.749,
+      "step": 2463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7467509201649687,
+      "learning_rate": 1.7553768085611033e-05,
+      "loss": 0.6228,
+      "step": 2464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6127320429072924,
+      "learning_rate": 1.7551608849815473e-05,
+      "loss": 0.796,
+      "step": 2465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4307340600762173,
+      "learning_rate": 1.754944879441087e-05,
+      "loss": 0.7548,
+      "step": 2466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7693917555197356,
+      "learning_rate": 1.7547287919631655e-05,
+      "loss": 0.8992,
+      "step": 2467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4706406131606666,
+      "learning_rate": 1.7545126225712366e-05,
+      "loss": 0.7496,
+      "step": 2468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6957896599223548,
+      "learning_rate": 1.7542963712887618e-05,
+      "loss": 0.6873,
+      "step": 2469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5260931443330856,
+      "learning_rate": 1.7540800381392116e-05,
+      "loss": 0.7522,
+      "step": 2470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.543857764791894,
+      "learning_rate": 1.753863623146066e-05,
+      "loss": 0.6656,
+      "step": 2471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4427949153903654,
+      "learning_rate": 1.753647126332813e-05,
+      "loss": 0.6484,
+      "step": 2472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.541434903268351,
+      "learning_rate": 1.7534305477229502e-05,
+      "loss": 0.6468,
+      "step": 2473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5649722223214777,
+      "learning_rate": 1.7532138873399838e-05,
+      "loss": 0.7066,
+      "step": 2474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.453940354232897,
+      "learning_rate": 1.7529971452074288e-05,
+      "loss": 0.7706,
+      "step": 2475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4539802973508147,
+      "learning_rate": 1.752780321348809e-05,
+      "loss": 0.635,
+      "step": 2476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5904603944217826,
+      "learning_rate": 1.752563415787658e-05,
+      "loss": 0.8014,
+      "step": 2477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6679468593874847,
+      "learning_rate": 1.752346428547517e-05,
+      "loss": 0.8044,
+      "step": 2478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5745163149114132,
+      "learning_rate": 1.7521293596519364e-05,
+      "loss": 0.835,
+      "step": 2479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6925493416084263,
+      "learning_rate": 1.7519122091244757e-05,
+      "loss": 0.7898,
+      "step": 2480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.511155784957144,
+      "learning_rate": 1.7516949769887032e-05,
+      "loss": 0.7317,
+      "step": 2481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6708420971092535,
+      "learning_rate": 1.7514776632681964e-05,
+      "loss": 0.8107,
+      "step": 2482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6091074751165642,
+      "learning_rate": 1.751260267986541e-05,
+      "loss": 0.7963,
+      "step": 2483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6040076390401958,
+      "learning_rate": 1.751042791167332e-05,
+      "loss": 0.7685,
+      "step": 2484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6669016961142142,
+      "learning_rate": 1.7508252328341726e-05,
+      "loss": 0.8528,
+      "step": 2485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5555218509357984,
+      "learning_rate": 1.7506075930106757e-05,
+      "loss": 0.7461,
+      "step": 2486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5085029479729102,
+      "learning_rate": 1.750389871720463e-05,
+      "loss": 0.8095,
+      "step": 2487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.647440048436809,
+      "learning_rate": 1.750172068987165e-05,
+      "loss": 0.8776,
+      "step": 2488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5388843511885144,
+      "learning_rate": 1.7499541848344198e-05,
+      "loss": 0.7754,
+      "step": 2489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.608332838142146,
+      "learning_rate": 1.749736219285876e-05,
+      "loss": 0.8276,
+      "step": 2490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4894710731185878,
+      "learning_rate": 1.7495181723651898e-05,
+      "loss": 0.7,
+      "step": 2491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4209838347335078,
+      "learning_rate": 1.7493000440960277e-05,
+      "loss": 0.6782,
+      "step": 2492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.506093175873103,
+      "learning_rate": 1.749081834502063e-05,
+      "loss": 0.6989,
+      "step": 2493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5791699181274723,
+      "learning_rate": 1.74886354360698e-05,
+      "loss": 0.7106,
+      "step": 2494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5764451475291832,
+      "learning_rate": 1.74864517143447e-05,
+      "loss": 0.7963,
+      "step": 2495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5289101233823514,
+      "learning_rate": 1.7484267180082343e-05,
+      "loss": 0.7024,
+      "step": 2496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.605164422817265,
+      "learning_rate": 1.7482081833519822e-05,
+      "loss": 0.847,
+      "step": 2497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.730092743675462,
+      "learning_rate": 1.7479895674894326e-05,
+      "loss": 0.6953,
+      "step": 2498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6454716877649915,
+      "learning_rate": 1.7477708704443125e-05,
+      "loss": 0.7739,
+      "step": 2499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8375789056356429,
+      "learning_rate": 1.7475520922403582e-05,
+      "loss": 0.877,
+      "step": 2500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5146743080412426,
+      "learning_rate": 1.7473332329013152e-05,
+      "loss": 0.7364,
+      "step": 2501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4101525134673571,
+      "learning_rate": 1.7471142924509365e-05,
+      "loss": 0.6998,
+      "step": 2502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4413863461289211,
+      "learning_rate": 1.7468952709129848e-05,
+      "loss": 0.7836,
+      "step": 2503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4701548661669526,
+      "learning_rate": 1.7466761683112316e-05,
+      "loss": 0.6526,
+      "step": 2504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.529401860380887,
+      "learning_rate": 1.746456984669457e-05,
+      "loss": 0.7717,
+      "step": 2505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3710047924443467,
+      "learning_rate": 1.7462377200114503e-05,
+      "loss": 0.6129,
+      "step": 2506
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7241252121013668,
+      "learning_rate": 1.7460183743610085e-05,
+      "loss": 0.7578,
+      "step": 2507
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6915304109956375,
+      "learning_rate": 1.745798947741939e-05,
+      "loss": 0.7747,
+      "step": 2508
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5089746597984957,
+      "learning_rate": 1.7455794401780565e-05,
+      "loss": 0.7323,
+      "step": 2509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4422751178761635,
+      "learning_rate": 1.745359851693185e-05,
+      "loss": 0.7499,
+      "step": 2510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5399503613199372,
+      "learning_rate": 1.7451401823111585e-05,
+      "loss": 0.7636,
+      "step": 2511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.606901545574598,
+      "learning_rate": 1.7449204320558172e-05,
+      "loss": 0.7686,
+      "step": 2512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5511141257902685,
+      "learning_rate": 1.7447006009510127e-05,
+      "loss": 0.6931,
+      "step": 2513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.425351819376338,
+      "learning_rate": 1.7444806890206036e-05,
+      "loss": 0.7941,
+      "step": 2514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3726498678601882,
+      "learning_rate": 1.7442606962884582e-05,
+      "loss": 0.7392,
+      "step": 2515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.430601606264957,
+      "learning_rate": 1.744040622778453e-05,
+      "loss": 0.6102,
+      "step": 2516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5338334677524086,
+      "learning_rate": 1.7438204685144733e-05,
+      "loss": 0.8332,
+      "step": 2517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5960453889328998,
+      "learning_rate": 1.7436002335204144e-05,
+      "loss": 0.8608,
+      "step": 2518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7945166823710936,
+      "learning_rate": 1.7433799178201786e-05,
+      "loss": 0.8269,
+      "step": 2519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.630442107761875,
+      "learning_rate": 1.743159521437678e-05,
+      "loss": 0.8059,
+      "step": 2520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5696632337208958,
+      "learning_rate": 1.7429390443968327e-05,
+      "loss": 0.847,
+      "step": 2521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7247206084773075,
+      "learning_rate": 1.742718486721573e-05,
+      "loss": 0.868,
+      "step": 2522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6226798368416193,
+      "learning_rate": 1.742497848435836e-05,
+      "loss": 0.8623,
+      "step": 2523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5800909522421087,
+      "learning_rate": 1.742277129563569e-05,
+      "loss": 0.7399,
+      "step": 2524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6787372496634005,
+      "learning_rate": 1.742056330128727e-05,
+      "loss": 0.6171,
+      "step": 2525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6313396899700607,
+      "learning_rate": 1.7418354501552755e-05,
+      "loss": 0.8866,
+      "step": 2526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6581527447564648,
+      "learning_rate": 1.7416144896671868e-05,
+      "loss": 0.7926,
+      "step": 2527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4389725304207477,
+      "learning_rate": 1.7413934486884424e-05,
+      "loss": 0.751,
+      "step": 2528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.626069977682484,
+      "learning_rate": 1.7411723272430337e-05,
+      "loss": 0.7008,
+      "step": 2529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5656126801302932,
+      "learning_rate": 1.7409511253549592e-05,
+      "loss": 0.7889,
+      "step": 2530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5549251618392606,
+      "learning_rate": 1.7407298430482272e-05,
+      "loss": 0.702,
+      "step": 2531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7298135822967657,
+      "learning_rate": 1.7405084803468545e-05,
+      "loss": 0.7869,
+      "step": 2532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5801692734885895,
+      "learning_rate": 1.7402870372748667e-05,
+      "loss": 0.7014,
+      "step": 2533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.585716907263995,
+      "learning_rate": 1.7400655138562977e-05,
+      "loss": 0.7708,
+      "step": 2534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6201214371053156,
+      "learning_rate": 1.7398439101151908e-05,
+      "loss": 0.7475,
+      "step": 2535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5630473760032266,
+      "learning_rate": 1.7396222260755967e-05,
+      "loss": 0.7259,
+      "step": 2536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6177675004741257,
+      "learning_rate": 1.739400461761577e-05,
+      "loss": 0.8223,
+      "step": 2537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5956924130781553,
+      "learning_rate": 1.7391786171971995e-05,
+      "loss": 0.8004,
+      "step": 2538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.548599637727936,
+      "learning_rate": 1.738956692406543e-05,
+      "loss": 0.7373,
+      "step": 2539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8006817203863534,
+      "learning_rate": 1.7387346874136936e-05,
+      "loss": 0.8084,
+      "step": 2540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5053069784925723,
+      "learning_rate": 1.7385126022427462e-05,
+      "loss": 0.7325,
+      "step": 2541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6085547890387042,
+      "learning_rate": 1.738290436917805e-05,
+      "loss": 0.8145,
+      "step": 2542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5814550489473806,
+      "learning_rate": 1.738068191462982e-05,
+      "loss": 0.7558,
+      "step": 2543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5549657373912378,
+      "learning_rate": 1.7378458659023995e-05,
+      "loss": 0.7423,
+      "step": 2544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5182984340635106,
+      "learning_rate": 1.7376234602601862e-05,
+      "loss": 0.7716,
+      "step": 2545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6084968643569708,
+      "learning_rate": 1.737400974560482e-05,
+      "loss": 0.8375,
+      "step": 2546
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.562666012399755,
+      "learning_rate": 1.7371784088274336e-05,
+      "loss": 0.7752,
+      "step": 2547
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4105680133899585,
+      "learning_rate": 1.7369557630851968e-05,
+      "loss": 0.7848,
+      "step": 2548
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6011744787383677,
+      "learning_rate": 1.736733037357937e-05,
+      "loss": 0.8372,
+      "step": 2549
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5803871758323422,
+      "learning_rate": 1.736510231669827e-05,
+      "loss": 0.7333,
+      "step": 2550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5883478183700042,
+      "learning_rate": 1.736287346045049e-05,
+      "loss": 0.8632,
+      "step": 2551
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6146855441975676,
+      "learning_rate": 1.736064380507794e-05,
+      "loss": 0.7427,
+      "step": 2552
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8021113066776384,
+      "learning_rate": 1.735841335082261e-05,
+      "loss": 0.8252,
+      "step": 2553
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5866143850955474,
+      "learning_rate": 1.7356182097926586e-05,
+      "loss": 0.7464,
+      "step": 2554
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4130768308137258,
+      "learning_rate": 1.7353950046632034e-05,
+      "loss": 0.7745,
+      "step": 2555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.707661990330429,
+      "learning_rate": 1.7351717197181205e-05,
+      "loss": 0.7891,
+      "step": 2556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7951534499962143,
+      "learning_rate": 1.7349483549816442e-05,
+      "loss": 0.7286,
+      "step": 2557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.499878405477149,
+      "learning_rate": 1.7347249104780174e-05,
+      "loss": 0.6814,
+      "step": 2558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.441729542410146,
+      "learning_rate": 1.7345013862314915e-05,
+      "loss": 0.7117,
+      "step": 2559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.485469000960217,
+      "learning_rate": 1.734277782266326e-05,
+      "loss": 0.7257,
+      "step": 2560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6231370649812686,
+      "learning_rate": 1.7340540986067907e-05,
+      "loss": 0.8767,
+      "step": 2561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5556532622432997,
+      "learning_rate": 1.733830335277162e-05,
+      "loss": 0.6962,
+      "step": 2562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5465549577351825,
+      "learning_rate": 1.733606492301726e-05,
+      "loss": 0.6955,
+      "step": 2563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5457692336086464,
+      "learning_rate": 1.7333825697047778e-05,
+      "loss": 0.7819,
+      "step": 2564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5236657144163215,
+      "learning_rate": 1.7331585675106205e-05,
+      "loss": 0.8064,
+      "step": 2565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.482979384172734,
+      "learning_rate": 1.732934485743566e-05,
+      "loss": 0.6854,
+      "step": 2566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5310665557094747,
+      "learning_rate": 1.732710324427935e-05,
+      "loss": 0.7453,
+      "step": 2567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5265915332766808,
+      "learning_rate": 1.7324860835880564e-05,
+      "loss": 0.7083,
+      "step": 2568
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5640457528269724,
+      "learning_rate": 1.7322617632482682e-05,
+      "loss": 0.8216,
+      "step": 2569
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5862358089141146,
+      "learning_rate": 1.732037363432917e-05,
+      "loss": 0.7527,
+      "step": 2570
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5229343485201696,
+      "learning_rate": 1.731812884166358e-05,
+      "loss": 0.7896,
+      "step": 2571
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4177542192456623,
+      "learning_rate": 1.7315883254729545e-05,
+      "loss": 0.7879,
+      "step": 2572
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6326061490316859,
+      "learning_rate": 1.731363687377079e-05,
+      "loss": 0.7037,
+      "step": 2573
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.601629614231217,
+      "learning_rate": 1.7311389699031123e-05,
+      "loss": 0.6777,
+      "step": 2574
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6520831590575529,
+      "learning_rate": 1.7309141730754445e-05,
+      "loss": 0.7052,
+      "step": 2575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.619518747814848,
+      "learning_rate": 1.730689296918473e-05,
+      "loss": 0.7202,
+      "step": 2576
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4296269965485597,
+      "learning_rate": 1.7304643414566054e-05,
+      "loss": 0.6845,
+      "step": 2577
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.489642595553068,
+      "learning_rate": 1.7302393067142567e-05,
+      "loss": 0.7439,
+      "step": 2578
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5089579003609654,
+      "learning_rate": 1.7300141927158506e-05,
+      "loss": 0.6531,
+      "step": 2579
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5543973690671014,
+      "learning_rate": 1.7297889994858207e-05,
+      "loss": 0.7204,
+      "step": 2580
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5813435578709207,
+      "learning_rate": 1.729563727048607e-05,
+      "loss": 0.7328,
+      "step": 2581
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.79196216418814,
+      "learning_rate": 1.7293383754286597e-05,
+      "loss": 0.8422,
+      "step": 2582
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4605108163184226,
+      "learning_rate": 1.7291129446504376e-05,
+      "loss": 0.7435,
+      "step": 2583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5836072342624712,
+      "learning_rate": 1.7288874347384074e-05,
+      "loss": 0.7172,
+      "step": 2584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5132036770627308,
+      "learning_rate": 1.7286618457170445e-05,
+      "loss": 0.733,
+      "step": 2585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4698388154169335,
+      "learning_rate": 1.7284361776108334e-05,
+      "loss": 0.7593,
+      "step": 2586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6601052501100197,
+      "learning_rate": 1.7282104304442665e-05,
+      "loss": 0.7853,
+      "step": 2587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6290132199377816,
+      "learning_rate": 1.7279846042418457e-05,
+      "loss": 0.8818,
+      "step": 2588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5030303854427125,
+      "learning_rate": 1.72775869902808e-05,
+      "loss": 0.7677,
+      "step": 2589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7115445257183575,
+      "learning_rate": 1.727532714827489e-05,
+      "loss": 0.7636,
+      "step": 2590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6349488084588633,
+      "learning_rate": 1.727306651664599e-05,
+      "loss": 0.85,
+      "step": 2591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5229765706901655,
+      "learning_rate": 1.7270805095639453e-05,
+      "loss": 0.7667,
+      "step": 2592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4680567877934216,
+      "learning_rate": 1.726854288550073e-05,
+      "loss": 0.741,
+      "step": 2593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5454570479760592,
+      "learning_rate": 1.726627988647534e-05,
+      "loss": 0.7619,
+      "step": 2594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4331477716947236,
+      "learning_rate": 1.7264016098808904e-05,
+      "loss": 0.7499,
+      "step": 2595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6029816212533754,
+      "learning_rate": 1.7261751522747116e-05,
+      "loss": 0.735,
+      "step": 2596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6261320064664497,
+      "learning_rate": 1.725948615853576e-05,
+      "loss": 0.8171,
+      "step": 2597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5537449429626073,
+      "learning_rate": 1.725722000642071e-05,
+      "loss": 0.7344,
+      "step": 2598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7498220500226265,
+      "learning_rate": 1.7254953066647915e-05,
+      "loss": 0.8478,
+      "step": 2599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4349222353392714,
+      "learning_rate": 1.7252685339463423e-05,
+      "loss": 0.7856,
+      "step": 2600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5870205146425946,
+      "learning_rate": 1.7250416825113355e-05,
+      "loss": 0.7084,
+      "step": 2601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5105277145464684,
+      "learning_rate": 1.7248147523843925e-05,
+      "loss": 0.7371,
+      "step": 2602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5539051554492644,
+      "learning_rate": 1.7245877435901428e-05,
+      "loss": 0.7413,
+      "step": 2603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5027853678248158,
+      "learning_rate": 1.7243606561532247e-05,
+      "loss": 0.8116,
+      "step": 2604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6730095843999155,
+      "learning_rate": 1.7241334900982854e-05,
+      "loss": 0.7399,
+      "step": 2605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.582166574687787,
+      "learning_rate": 1.72390624544998e-05,
+      "loss": 0.7094,
+      "step": 2606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6502380774099639,
+      "learning_rate": 1.723678922232972e-05,
+      "loss": 0.7626,
+      "step": 2607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.666325386956416,
+      "learning_rate": 1.7234515204719342e-05,
+      "loss": 0.7526,
+      "step": 2608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5806104358843491,
+      "learning_rate": 1.7232240401915473e-05,
+      "loss": 0.7783,
+      "step": 2609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5524286203165254,
+      "learning_rate": 1.722996481416501e-05,
+      "loss": 0.7713,
+      "step": 2610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5354166257297275,
+      "learning_rate": 1.722768844171493e-05,
+      "loss": 0.7797,
+      "step": 2611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.648540577003657,
+      "learning_rate": 1.7225411284812294e-05,
+      "loss": 0.8315,
+      "step": 2612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4579288968385267,
+      "learning_rate": 1.722313334370426e-05,
+      "loss": 0.8666,
+      "step": 2613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4702550601601378,
+      "learning_rate": 1.722085461863806e-05,
+      "loss": 0.6591,
+      "step": 2614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6201922996801192,
+      "learning_rate": 1.721857510986101e-05,
+      "loss": 0.7371,
+      "step": 2615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5990364634319834,
+      "learning_rate": 1.7216294817620518e-05,
+      "loss": 0.8072,
+      "step": 2616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5437191298534145,
+      "learning_rate": 1.7214013742164072e-05,
+      "loss": 0.846,
+      "step": 2617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5671733942399428,
+      "learning_rate": 1.7211731883739254e-05,
+      "loss": 0.7216,
+      "step": 2618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5207270159074848,
+      "learning_rate": 1.7209449242593715e-05,
+      "loss": 0.8302,
+      "step": 2619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4966851389973288,
+      "learning_rate": 1.7207165818975206e-05,
+      "loss": 0.7077,
+      "step": 2620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6020631870179807,
+      "learning_rate": 1.720488161313155e-05,
+      "loss": 0.7177,
+      "step": 2621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6527642932203361,
+      "learning_rate": 1.720259662531067e-05,
+      "loss": 0.8524,
+      "step": 2622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5292848541794821,
+      "learning_rate": 1.7200310855760564e-05,
+      "loss": 0.6932,
+      "step": 2623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5040037411224556,
+      "learning_rate": 1.7198024304729314e-05,
+      "loss": 0.7989,
+      "step": 2624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6330275027530194,
+      "learning_rate": 1.7195736972465087e-05,
+      "loss": 0.7362,
+      "step": 2625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4973184520025868,
+      "learning_rate": 1.7193448859216146e-05,
+      "loss": 0.7976,
+      "step": 2626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5210898000562127,
+      "learning_rate": 1.719115996523082e-05,
+      "loss": 0.8209,
+      "step": 2627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.823776645172069,
+      "learning_rate": 1.7188870290757536e-05,
+      "loss": 0.6101,
+      "step": 2628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3714876037564456,
+      "learning_rate": 1.7186579836044804e-05,
+      "loss": 0.7237,
+      "step": 2629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5517520331930001,
+      "learning_rate": 1.718428860134122e-05,
+      "loss": 0.7674,
+      "step": 2630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5261112683689515,
+      "learning_rate": 1.7181996586895456e-05,
+      "loss": 0.7371,
+      "step": 2631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.546182807578598,
+      "learning_rate": 1.7179703792956276e-05,
+      "loss": 0.6897,
+      "step": 2632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6211807619267422,
+      "learning_rate": 1.7177410219772527e-05,
+      "loss": 0.7503,
+      "step": 2633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6640469296407736,
+      "learning_rate": 1.7175115867593143e-05,
+      "loss": 0.8524,
+      "step": 2634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5480571031288723,
+      "learning_rate": 1.7172820736667133e-05,
+      "loss": 0.7468,
+      "step": 2635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.685157331961948,
+      "learning_rate": 1.7170524827243608e-05,
+      "loss": 0.7039,
+      "step": 2636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5492580721226499,
+      "learning_rate": 1.7168228139571744e-05,
+      "loss": 0.8109,
+      "step": 2637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4887599314971207,
+      "learning_rate": 1.7165930673900812e-05,
+      "loss": 0.7612,
+      "step": 2638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6865716116863976,
+      "learning_rate": 1.7163632430480172e-05,
+      "loss": 0.7688,
+      "step": 2639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6204207493635825,
+      "learning_rate": 1.7161333409559256e-05,
+      "loss": 0.7999,
+      "step": 2640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5023197048656975,
+      "learning_rate": 1.715903361138759e-05,
+      "loss": 0.7147,
+      "step": 2641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4427185908059887,
+      "learning_rate": 1.715673303621478e-05,
+      "loss": 0.6728,
+      "step": 2642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4589691465012895,
+      "learning_rate": 1.715443168429052e-05,
+      "loss": 0.8163,
+      "step": 2643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.622314886789072,
+      "learning_rate": 1.7152129555864583e-05,
+      "loss": 0.7668,
+      "step": 2644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5875141396662342,
+      "learning_rate": 1.7149826651186828e-05,
+      "loss": 0.7138,
+      "step": 2645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5819705613379111,
+      "learning_rate": 1.71475229705072e-05,
+      "loss": 0.7547,
+      "step": 2646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5286856655549894,
+      "learning_rate": 1.714521851407573e-05,
+      "loss": 0.6983,
+      "step": 2647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6680424585241245,
+      "learning_rate": 1.7142913282142528e-05,
+      "loss": 0.8304,
+      "step": 2648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3949753968844358,
+      "learning_rate": 1.7140607274957793e-05,
+      "loss": 0.7176,
+      "step": 2649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7513081690313692,
+      "learning_rate": 1.7138300492771803e-05,
+      "loss": 0.7674,
+      "step": 2650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5134706541956857,
+      "learning_rate": 1.7135992935834927e-05,
+      "loss": 0.6783,
+      "step": 2651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.749163817172862,
+      "learning_rate": 1.7133684604397613e-05,
+      "loss": 0.7608,
+      "step": 2652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4252020954515696,
+      "learning_rate": 1.713137549871039e-05,
+      "loss": 0.764,
+      "step": 2653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6358995582805185,
+      "learning_rate": 1.712906561902388e-05,
+      "loss": 0.7249,
+      "step": 2654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.535740650980999,
+      "learning_rate": 1.7126754965588785e-05,
+      "loss": 0.6888,
+      "step": 2655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2716763178954538,
+      "learning_rate": 1.7124443538655887e-05,
+      "loss": 0.6396,
+      "step": 2656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5575463059511079,
+      "learning_rate": 1.7122131338476058e-05,
+      "loss": 0.6889,
+      "step": 2657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5997207855607138,
+      "learning_rate": 1.7119818365300246e-05,
+      "loss": 0.735,
+      "step": 2658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.65220121537614,
+      "learning_rate": 1.7117504619379498e-05,
+      "loss": 0.7919,
+      "step": 2659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5541586302274732,
+      "learning_rate": 1.7115190100964926e-05,
+      "loss": 0.7166,
+      "step": 2660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4452623372732751,
+      "learning_rate": 1.7112874810307736e-05,
+      "loss": 0.68,
+      "step": 2661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5159176357495028,
+      "learning_rate": 1.711055874765922e-05,
+      "loss": 0.8157,
+      "step": 2662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5363560553715614,
+      "learning_rate": 1.710824191327075e-05,
+      "loss": 0.7492,
+      "step": 2663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7141688792271914,
+      "learning_rate": 1.7105924307393776e-05,
+      "loss": 0.7959,
+      "step": 2664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6963032092016188,
+      "learning_rate": 1.7103605930279847e-05,
+      "loss": 0.789,
+      "step": 2665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.570563090841888,
+      "learning_rate": 1.7101286782180585e-05,
+      "loss": 0.7145,
+      "step": 2666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5015102151992468,
+      "learning_rate": 1.7098966863347687e-05,
+      "loss": 0.7126,
+      "step": 2667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.523283582923896,
+      "learning_rate": 1.7096646174032955e-05,
+      "loss": 0.7493,
+      "step": 2668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6155367999750465,
+      "learning_rate": 1.7094324714488258e-05,
+      "loss": 0.7459,
+      "step": 2669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4925118200717544,
+      "learning_rate": 1.709200248496556e-05,
+      "loss": 0.8062,
+      "step": 2670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4778170137928905,
+      "learning_rate": 1.7089679485716897e-05,
+      "loss": 0.7196,
+      "step": 2671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4904397859832723,
+      "learning_rate": 1.7087355716994398e-05,
+      "loss": 0.6981,
+      "step": 2672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7317550031169902,
+      "learning_rate": 1.7085031179050268e-05,
+      "loss": 0.7211,
+      "step": 2673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.489607713671086,
+      "learning_rate": 1.7082705872136797e-05,
+      "loss": 0.7409,
+      "step": 2674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.516574391694995,
+      "learning_rate": 1.708037979650637e-05,
+      "loss": 0.7761,
+      "step": 2675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6355128601376654,
+      "learning_rate": 1.707805295241144e-05,
+      "loss": 0.8915,
+      "step": 2676
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5304123605760016,
+      "learning_rate": 1.707572534010455e-05,
+      "loss": 0.849,
+      "step": 2677
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5575421176033872,
+      "learning_rate": 1.7073396959838326e-05,
+      "loss": 0.724,
+      "step": 2678
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.696531813138858,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.7578,
+      "step": 2679
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6790816568746896,
+      "learning_rate": 1.7068737896438796e-05,
+      "loss": 0.8616,
+      "step": 2680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7031711209522635,
+      "learning_rate": 1.706640721381116e-05,
+      "loss": 0.7563,
+      "step": 2681
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4519377943836558,
+      "learning_rate": 1.7064075764235527e-05,
+      "loss": 0.6649,
+      "step": 2682
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6649570253386337,
+      "learning_rate": 1.706174354796494e-05,
+      "loss": 0.8154,
+      "step": 2683
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.620028413852496,
+      "learning_rate": 1.7059410565252525e-05,
+      "loss": 0.7565,
+      "step": 2684
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.355628318623081,
+      "learning_rate": 1.7057076816351487e-05,
+      "loss": 0.5766,
+      "step": 2685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4905400260550263,
+      "learning_rate": 1.7054742301515123e-05,
+      "loss": 0.6522,
+      "step": 2686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6314227432654287,
+      "learning_rate": 1.7052407020996804e-05,
+      "loss": 0.7687,
+      "step": 2687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6353452003075364,
+      "learning_rate": 1.7050070975049993e-05,
+      "loss": 0.7895,
+      "step": 2688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3951006709312348,
+      "learning_rate": 1.7047734163928227e-05,
+      "loss": 0.7829,
+      "step": 2689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6440774612884574,
+      "learning_rate": 1.704539658788513e-05,
+      "loss": 0.8283,
+      "step": 2690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4537194274268685,
+      "learning_rate": 1.7043058247174414e-05,
+      "loss": 0.639,
+      "step": 2691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6083947454220908,
+      "learning_rate": 1.7040719142049864e-05,
+      "loss": 0.6641,
+      "step": 2692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6615547545200424,
+      "learning_rate": 1.7038379272765357e-05,
+      "loss": 0.7293,
+      "step": 2693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5466180268091978,
+      "learning_rate": 1.703603863957485e-05,
+      "loss": 0.7146,
+      "step": 2694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5668440062466291,
+      "learning_rate": 1.7033697242732376e-05,
+      "loss": 0.8263,
+      "step": 2695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7239135169274935,
+      "learning_rate": 1.7031355082492066e-05,
+      "loss": 0.8717,
+      "step": 2696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.484778336143205,
+      "learning_rate": 1.7029012159108114e-05,
+      "loss": 0.7309,
+      "step": 2697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.550338169624176,
+      "learning_rate": 1.7026668472834818e-05,
+      "loss": 0.7148,
+      "step": 2698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4474445970065117,
+      "learning_rate": 1.702432402392654e-05,
+      "loss": 0.7862,
+      "step": 2699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.61847983913502,
+      "learning_rate": 1.7021978812637742e-05,
+      "loss": 0.7613,
+      "step": 2700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5619155946150487,
+      "learning_rate": 1.7019632839222954e-05,
+      "loss": 0.8718,
+      "step": 2701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5338774499992636,
+      "learning_rate": 1.7017286103936793e-05,
+      "loss": 0.8118,
+      "step": 2702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3798929312784296,
+      "learning_rate": 1.7014938607033967e-05,
+      "loss": 0.7618,
+      "step": 2703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6002922978466065,
+      "learning_rate": 1.7012590348769255e-05,
+      "loss": 0.7193,
+      "step": 2704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4831134418629623,
+      "learning_rate": 1.7010241329397524e-05,
+      "loss": 0.7847,
+      "step": 2705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4187030989217948,
+      "learning_rate": 1.700789154917373e-05,
+      "loss": 0.6872,
+      "step": 2706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5502049567007963,
+      "learning_rate": 1.7005541008352894e-05,
+      "loss": 0.6629,
+      "step": 2707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.562381202598881,
+      "learning_rate": 1.700318970719014e-05,
+      "loss": 0.7573,
+      "step": 2708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4838134273125427,
+      "learning_rate": 1.7000837645940654e-05,
+      "loss": 0.7006,
+      "step": 2709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.562643798332321,
+      "learning_rate": 1.6998484824859728e-05,
+      "loss": 0.8275,
+      "step": 2710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4447304429418308,
+      "learning_rate": 1.699613124420272e-05,
+      "loss": 0.7452,
+      "step": 2711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4520248132390094,
+      "learning_rate": 1.699377690422507e-05,
+      "loss": 0.794,
+      "step": 2712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5650615558288783,
+      "learning_rate": 1.6991421805182305e-05,
+      "loss": 0.8639,
+      "step": 2713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5385647957238642,
+      "learning_rate": 1.6989065947330038e-05,
+      "loss": 0.7816,
+      "step": 2714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.605033676576137,
+      "learning_rate": 1.6986709330923964e-05,
+      "loss": 0.7617,
+      "step": 2715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5254040652379102,
+      "learning_rate": 1.6984351956219847e-05,
+      "loss": 0.7744,
+      "step": 2716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5370041367141265,
+      "learning_rate": 1.698199382347355e-05,
+      "loss": 0.7817,
+      "step": 2717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5591871015222685,
+      "learning_rate": 1.697963493294101e-05,
+      "loss": 0.7608,
+      "step": 2718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5268100390967951,
+      "learning_rate": 1.6977275284878243e-05,
+      "loss": 0.7932,
+      "step": 2719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5310722382386455,
+      "learning_rate": 1.697491487954136e-05,
+      "loss": 0.7739,
+      "step": 2720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5020818896731902,
+      "learning_rate": 1.6972553717186543e-05,
+      "loss": 0.7654,
+      "step": 2721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6285379518225092,
+      "learning_rate": 1.6970191798070056e-05,
+      "loss": 0.8342,
+      "step": 2722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5214473900346936,
+      "learning_rate": 1.6967829122448255e-05,
+      "loss": 0.7637,
+      "step": 2723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6437973758314557,
+      "learning_rate": 1.6965465690577562e-05,
+      "loss": 0.6452,
+      "step": 2724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5907319901063022,
+      "learning_rate": 1.69631015027145e-05,
+      "loss": 0.7918,
+      "step": 2725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6153610345004459,
+      "learning_rate": 1.696073655911566e-05,
+      "loss": 0.6448,
+      "step": 2726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6265533051979708,
+      "learning_rate": 1.695837086003772e-05,
+      "loss": 0.7247,
+      "step": 2727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7536067425885253,
+      "learning_rate": 1.695600440573744e-05,
+      "loss": 0.8441,
+      "step": 2728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4307599485709788,
+      "learning_rate": 1.6953637196471658e-05,
+      "loss": 0.6191,
+      "step": 2729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.369399874389617,
+      "learning_rate": 1.6951269232497307e-05,
+      "loss": 0.7959,
+      "step": 2730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5347249065403918,
+      "learning_rate": 1.6948900514071383e-05,
+      "loss": 0.7254,
+      "step": 2731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.569237032596756,
+      "learning_rate": 1.6946531041450976e-05,
+      "loss": 0.843,
+      "step": 2732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4274347398074667,
+      "learning_rate": 1.694416081489326e-05,
+      "loss": 0.7344,
+      "step": 2733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5881966084737997,
+      "learning_rate": 1.694178983465548e-05,
+      "loss": 0.7619,
+      "step": 2734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4842518701692677,
+      "learning_rate": 1.6939418100994972e-05,
+      "loss": 0.6843,
+      "step": 2735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.409863595614964,
+      "learning_rate": 1.6937045614169147e-05,
+      "loss": 0.6616,
+      "step": 2736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7392852185664112,
+      "learning_rate": 1.6934672374435504e-05,
+      "loss": 0.7943,
+      "step": 2737
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6435124089413926,
+      "learning_rate": 1.693229838205162e-05,
+      "loss": 0.7713,
+      "step": 2738
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4599316988101887,
+      "learning_rate": 1.692992363727516e-05,
+      "loss": 0.8101,
+      "step": 2739
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.489997692489692,
+      "learning_rate": 1.6927548140363857e-05,
+      "loss": 0.7271,
+      "step": 2740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.62368929242668,
+      "learning_rate": 1.692517189157554e-05,
+      "loss": 0.8319,
+      "step": 2741
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6038961839172665,
+      "learning_rate": 1.6922794891168114e-05,
+      "loss": 0.6931,
+      "step": 2742
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6660621188481828,
+      "learning_rate": 1.6920417139399558e-05,
+      "loss": 0.7983,
+      "step": 2743
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6005730767318995,
+      "learning_rate": 1.6918038636527947e-05,
+      "loss": 0.7519,
+      "step": 2744
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5824965838797156,
+      "learning_rate": 1.691565938281143e-05,
+      "loss": 0.8057,
+      "step": 2745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6253194656473673,
+      "learning_rate": 1.691327937850823e-05,
+      "loss": 0.776,
+      "step": 2746
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5283801314123524,
+      "learning_rate": 1.691089862387667e-05,
+      "loss": 0.6774,
+      "step": 2747
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4364607171702077,
+      "learning_rate": 1.690851711917514e-05,
+      "loss": 0.7053,
+      "step": 2748
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4031659843312712,
+      "learning_rate": 1.690613486466211e-05,
+      "loss": 0.7023,
+      "step": 2749
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5841263671334511,
+      "learning_rate": 1.6903751860596142e-05,
+      "loss": 0.7127,
+      "step": 2750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6482604897965214,
+      "learning_rate": 1.690136810723587e-05,
+      "loss": 0.8546,
+      "step": 2751
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4450313742005596,
+      "learning_rate": 1.689898360484002e-05,
+      "loss": 0.8229,
+      "step": 2752
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6073793491502968,
+      "learning_rate": 1.6896598353667384e-05,
+      "loss": 0.7387,
+      "step": 2753
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7321541961311,
+      "learning_rate": 1.689421235397685e-05,
+      "loss": 0.7184,
+      "step": 2754
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.595811929094905,
+      "learning_rate": 1.6891825606027376e-05,
+      "loss": 0.8199,
+      "step": 2755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.653256316090929,
+      "learning_rate": 1.688943811007801e-05,
+      "loss": 0.7459,
+      "step": 2756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3943982750909278,
+      "learning_rate": 1.6887049866387874e-05,
+      "loss": 0.7861,
+      "step": 2757
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.833818403056107,
+      "learning_rate": 1.6884660875216178e-05,
+      "loss": 0.8094,
+      "step": 2758
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.62709212438958,
+      "learning_rate": 1.6882271136822204e-05,
+      "loss": 0.8247,
+      "step": 2759
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4804818862878864,
+      "learning_rate": 1.687988065146533e-05,
+      "loss": 0.7697,
+      "step": 2760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.535185488000713,
+      "learning_rate": 1.6877489419405e-05,
+      "loss": 0.7532,
+      "step": 2761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4658145091412107,
+      "learning_rate": 1.6875097440900746e-05,
+      "loss": 0.7117,
+      "step": 2762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.476554530620898,
+      "learning_rate": 1.6872704716212178e-05,
+      "loss": 0.696,
+      "step": 2763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4988545882011124,
+      "learning_rate": 1.6870311245598992e-05,
+      "loss": 0.8044,
+      "step": 2764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4127927069516595,
+      "learning_rate": 1.6867917029320958e-05,
+      "loss": 0.8519,
+      "step": 2765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3757450216152427,
+      "learning_rate": 1.6865522067637932e-05,
+      "loss": 0.6608,
+      "step": 2766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4416391583154726,
+      "learning_rate": 1.686312636080985e-05,
+      "loss": 0.6738,
+      "step": 2767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4930160707420024,
+      "learning_rate": 1.686072990909673e-05,
+      "loss": 0.7284,
+      "step": 2768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6919012095067527,
+      "learning_rate": 1.6858332712758667e-05,
+      "loss": 0.7883,
+      "step": 2769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4702667160247174,
+      "learning_rate": 1.6855934772055843e-05,
+      "loss": 0.7797,
+      "step": 2770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4828486258168854,
+      "learning_rate": 1.6853536087248512e-05,
+      "loss": 0.8487,
+      "step": 2771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6255330339109715,
+      "learning_rate": 1.6851136658597014e-05,
+      "loss": 0.7718,
+      "step": 2772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5772887860309583,
+      "learning_rate": 1.6848736486361777e-05,
+      "loss": 0.6948,
+      "step": 2773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7697523320924025,
+      "learning_rate": 1.6846335570803293e-05,
+      "loss": 0.8107,
+      "step": 2774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.575249049219454,
+      "learning_rate": 1.6843933912182148e-05,
+      "loss": 0.7582,
+      "step": 2775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6035336350257257,
+      "learning_rate": 1.6841531510759007e-05,
+      "loss": 0.7807,
+      "step": 2776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5277565031061031,
+      "learning_rate": 1.6839128366794603e-05,
+      "loss": 0.8003,
+      "step": 2777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.665939020688003,
+      "learning_rate": 1.6836724480549773e-05,
+      "loss": 0.7279,
+      "step": 2778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.649265018826581,
+      "learning_rate": 1.6834319852285413e-05,
+      "loss": 0.7026,
+      "step": 2779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5585497816292266,
+      "learning_rate": 1.6831914482262514e-05,
+      "loss": 0.7922,
+      "step": 2780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5025288970479587,
+      "learning_rate": 1.682950837074213e-05,
+      "loss": 0.6813,
+      "step": 2781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.703071819086007,
+      "learning_rate": 1.6827101517985417e-05,
+      "loss": 0.8524,
+      "step": 2782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.483095775477807,
+      "learning_rate": 1.6824693924253596e-05,
+      "loss": 0.6962,
+      "step": 2783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.48203059260217,
+      "learning_rate": 1.682228558980798e-05,
+      "loss": 0.768,
+      "step": 2784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6182378066924379,
+      "learning_rate": 1.681987651490995e-05,
+      "loss": 0.7338,
+      "step": 2785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5946567346322944,
+      "learning_rate": 1.6817466699820976e-05,
+      "loss": 0.7571,
+      "step": 2786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4595787577320416,
+      "learning_rate": 1.6815056144802604e-05,
+      "loss": 0.7257,
+      "step": 2787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7789185563151684,
+      "learning_rate": 1.681264485011646e-05,
+      "loss": 0.6834,
+      "step": 2788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5248080417971017,
+      "learning_rate": 1.681023281602426e-05,
+      "loss": 0.7149,
+      "step": 2789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.425715654156925,
+      "learning_rate": 1.6807820042787788e-05,
+      "loss": 0.7185,
+      "step": 2790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4767606646263185,
+      "learning_rate": 1.680540653066891e-05,
+      "loss": 0.8509,
+      "step": 2791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.694444133436615,
+      "learning_rate": 1.6802992279929583e-05,
+      "loss": 0.8359,
+      "step": 2792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6278746176394097,
+      "learning_rate": 1.6800577290831825e-05,
+      "loss": 0.698,
+      "step": 2793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4507274176318135,
+      "learning_rate": 1.6798161563637754e-05,
+      "loss": 0.689,
+      "step": 2794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4968155686337008,
+      "learning_rate": 1.6795745098609555e-05,
+      "loss": 0.6953,
+      "step": 2795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4824801508895227,
+      "learning_rate": 1.6793327896009498e-05,
+      "loss": 0.6898,
+      "step": 2796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6046740497702996,
+      "learning_rate": 1.6790909956099935e-05,
+      "loss": 0.6818,
+      "step": 2797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5459818527277547,
+      "learning_rate": 1.6788491279143298e-05,
+      "loss": 0.682,
+      "step": 2798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4972000594763024,
+      "learning_rate": 1.6786071865402086e-05,
+      "loss": 0.7576,
+      "step": 2799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.560378258512186,
+      "learning_rate": 1.6783651715138902e-05,
+      "loss": 0.7076,
+      "step": 2800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6391118154486723,
+      "learning_rate": 1.6781230828616404e-05,
+      "loss": 0.8201,
+      "step": 2801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6916916979306973,
+      "learning_rate": 1.6778809206097347e-05,
+      "loss": 0.8081,
+      "step": 2802
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6874205828232405,
+      "learning_rate": 1.6776386847844563e-05,
+      "loss": 0.7184,
+      "step": 2803
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.458018963147143,
+      "learning_rate": 1.6773963754120952e-05,
+      "loss": 0.7259,
+      "step": 2804
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6070999871208667,
+      "learning_rate": 1.6771539925189514e-05,
+      "loss": 0.7739,
+      "step": 2805
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.666769768470216,
+      "learning_rate": 1.6769115361313308e-05,
+      "loss": 0.7305,
+      "step": 2806
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.513886999979369,
+      "learning_rate": 1.6766690062755488e-05,
+      "loss": 0.6507,
+      "step": 2807
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5865848169441699,
+      "learning_rate": 1.676426402977928e-05,
+      "loss": 0.6846,
+      "step": 2808
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5473956402439422,
+      "learning_rate": 1.6761837262647994e-05,
+      "loss": 0.7615,
+      "step": 2809
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7061180332352512,
+      "learning_rate": 1.6759409761625015e-05,
+      "loss": 0.7268,
+      "step": 2810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5375806434832613,
+      "learning_rate": 1.6756981526973813e-05,
+      "loss": 0.7344,
+      "step": 2811
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.626081660308829,
+      "learning_rate": 1.6754552558957936e-05,
+      "loss": 0.7418,
+      "step": 2812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4755972980689747,
+      "learning_rate": 1.6752122857841005e-05,
+      "loss": 0.7499,
+      "step": 2813
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6415653421485688,
+      "learning_rate": 1.6749692423886727e-05,
+      "loss": 0.7489,
+      "step": 2814
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5273140977187238,
+      "learning_rate": 1.6747261257358894e-05,
+      "loss": 0.8002,
+      "step": 2815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5374915960526312,
+      "learning_rate": 1.6744829358521368e-05,
+      "loss": 0.7432,
+      "step": 2816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.488772374315502,
+      "learning_rate": 1.674239672763809e-05,
+      "loss": 0.6842,
+      "step": 2817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4171710004641913,
+      "learning_rate": 1.6739963364973084e-05,
+      "loss": 0.6569,
+      "step": 2818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6203595465668792,
+      "learning_rate": 1.673752927079046e-05,
+      "loss": 0.7785,
+      "step": 2819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.493613756350887,
+      "learning_rate": 1.6735094445354395e-05,
+      "loss": 0.8533,
+      "step": 2820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5122966135044509,
+      "learning_rate": 1.6732658888929153e-05,
+      "loss": 0.8206,
+      "step": 2821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4885803944048208,
+      "learning_rate": 1.6730222601779075e-05,
+      "loss": 0.6961,
+      "step": 2822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6084817399112654,
+      "learning_rate": 1.6727785584168583e-05,
+      "loss": 0.7868,
+      "step": 2823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5461636352471053,
+      "learning_rate": 1.6725347836362178e-05,
+      "loss": 0.6473,
+      "step": 2824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5190104114024647,
+      "learning_rate": 1.6722909358624436e-05,
+      "loss": 0.7424,
+      "step": 2825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5882905731886237,
+      "learning_rate": 1.6720470151220018e-05,
+      "loss": 0.7879,
+      "step": 2826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6363752670735936,
+      "learning_rate": 1.671803021441366e-05,
+      "loss": 0.7221,
+      "step": 2827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7099595353026695,
+      "learning_rate": 1.6715589548470187e-05,
+      "loss": 0.8382,
+      "step": 2828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6410804954346245,
+      "learning_rate": 1.6713148153654483e-05,
+      "loss": 0.8053,
+      "step": 2829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.440973928004605,
+      "learning_rate": 1.6710706030231533e-05,
+      "loss": 0.733,
+      "step": 2830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.45683336566356,
+      "learning_rate": 1.670826317846638e-05,
+      "loss": 0.7442,
+      "step": 2831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5222653629474279,
+      "learning_rate": 1.6705819598624168e-05,
+      "loss": 0.7067,
+      "step": 2832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5092845480116759,
+      "learning_rate": 1.6703375290970107e-05,
+      "loss": 0.6241,
+      "step": 2833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.673406426838146,
+      "learning_rate": 1.6700930255769486e-05,
+      "loss": 0.9131,
+      "step": 2834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4722136796614742,
+      "learning_rate": 1.6698484493287678e-05,
+      "loss": 0.774,
+      "step": 2835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7138172095048119,
+      "learning_rate": 1.6696038003790128e-05,
+      "loss": 0.8088,
+      "step": 2836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4999575241202465,
+      "learning_rate": 1.6693590787542372e-05,
+      "loss": 0.8001,
+      "step": 2837
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.596426220710093,
+      "learning_rate": 1.6691142844810013e-05,
+      "loss": 0.7445,
+      "step": 2838
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5321179361818746,
+      "learning_rate": 1.668869417585873e-05,
+      "loss": 0.7331,
+      "step": 2839
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4175973021799633,
+      "learning_rate": 1.6686244780954294e-05,
+      "loss": 0.6561,
+      "step": 2840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6761302463380996,
+      "learning_rate": 1.668379466036255e-05,
+      "loss": 0.7373,
+      "step": 2841
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6008508892660749,
+      "learning_rate": 1.668134381434942e-05,
+      "loss": 0.7636,
+      "step": 2842
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7527217040594256,
+      "learning_rate": 1.66788922431809e-05,
+      "loss": 0.8217,
+      "step": 2843
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.49467108389566,
+      "learning_rate": 1.6676439947123075e-05,
+      "loss": 0.7853,
+      "step": 2844
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6393797745455427,
+      "learning_rate": 1.66739869264421e-05,
+      "loss": 0.8308,
+      "step": 2845
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4602628736605359,
+      "learning_rate": 1.667153318140422e-05,
+      "loss": 0.7054,
+      "step": 2846
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.579963462481767,
+      "learning_rate": 1.666907871227574e-05,
+      "loss": 0.8356,
+      "step": 2847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6868043052083224,
+      "learning_rate": 1.6666623519323056e-05,
+      "loss": 0.7385,
+      "step": 2848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5060864355496348,
+      "learning_rate": 1.6664167602812646e-05,
+      "loss": 0.7126,
+      "step": 2849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.369964772802513,
+      "learning_rate": 1.6661710963011057e-05,
+      "loss": 0.7056,
+      "step": 2850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5409934582076479,
+      "learning_rate": 1.665925360018492e-05,
+      "loss": 0.7766,
+      "step": 2851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5099368755717504,
+      "learning_rate": 1.6656795514600948e-05,
+      "loss": 0.8421,
+      "step": 2852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4299528658124159,
+      "learning_rate": 1.665433670652592e-05,
+      "loss": 0.7769,
+      "step": 2853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.519976430392634,
+      "learning_rate": 1.665187717622671e-05,
+      "loss": 0.7765,
+      "step": 2854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4572693694812913,
+      "learning_rate": 1.6649416923970248e-05,
+      "loss": 0.7355,
+      "step": 2855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.630645817094188,
+      "learning_rate": 1.664695595002357e-05,
+      "loss": 0.7639,
+      "step": 2856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7757479522459132,
+      "learning_rate": 1.6644494254653767e-05,
+      "loss": 0.8451,
+      "step": 2857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.558065617309748,
+      "learning_rate": 1.664203183812802e-05,
+      "loss": 0.774,
+      "step": 2858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5455113506164675,
+      "learning_rate": 1.6639568700713587e-05,
+      "loss": 0.7113,
+      "step": 2859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.699674973561388,
+      "learning_rate": 1.6637104842677807e-05,
+      "loss": 0.6718,
+      "step": 2860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5690022382016071,
+      "learning_rate": 1.6634640264288087e-05,
+      "loss": 0.8394,
+      "step": 2861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5342184544437765,
+      "learning_rate": 1.663217496581192e-05,
+      "loss": 0.795,
+      "step": 2862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5959621667132664,
+      "learning_rate": 1.6629708947516875e-05,
+      "loss": 0.8409,
+      "step": 2863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3836758638315834,
+      "learning_rate": 1.6627242209670606e-05,
+      "loss": 0.7884,
+      "step": 2864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7082254637709893,
+      "learning_rate": 1.6624774752540828e-05,
+      "loss": 0.848,
+      "step": 2865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4559456057698992,
+      "learning_rate": 1.6622306576395355e-05,
+      "loss": 0.7191,
+      "step": 2866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.491862156638078,
+      "learning_rate": 1.661983768150206e-05,
+      "loss": 0.779,
+      "step": 2867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.58902897489258,
+      "learning_rate": 1.661736806812891e-05,
+      "loss": 0.7634,
+      "step": 2868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.452695928938561,
+      "learning_rate": 1.661489773654394e-05,
+      "loss": 0.6678,
+      "step": 2869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4349696118715132,
+      "learning_rate": 1.6612426687015266e-05,
+      "loss": 0.7108,
+      "step": 2870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.578562608016945,
+      "learning_rate": 1.6609954919811078e-05,
+      "loss": 0.7473,
+      "step": 2871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5552003153913527,
+      "learning_rate": 1.6607482435199654e-05,
+      "loss": 0.7015,
+      "step": 2872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5501887712427833,
+      "learning_rate": 1.6605009233449344e-05,
+      "loss": 0.8025,
+      "step": 2873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5822623419828576,
+      "learning_rate": 1.6602535314828572e-05,
+      "loss": 0.7478,
+      "step": 2874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.563036489716803,
+      "learning_rate": 1.6600060679605844e-05,
+      "loss": 0.7182,
+      "step": 2875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5476161809098843,
+      "learning_rate": 1.6597585328049736e-05,
+      "loss": 0.6821,
+      "step": 2876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.496579865043439,
+      "learning_rate": 1.6595109260428924e-05,
+      "loss": 0.6769,
+      "step": 2877
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.595038129995531,
+      "learning_rate": 1.6592632477012132e-05,
+      "loss": 0.8206,
+      "step": 2878
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6403593358151038,
+      "learning_rate": 1.6590154978068185e-05,
+      "loss": 0.7059,
+      "step": 2879
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4723084470416206,
+      "learning_rate": 1.6587676763865972e-05,
+      "loss": 0.7159,
+      "step": 2880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4455832157117119,
+      "learning_rate": 1.6585197834674466e-05,
+      "loss": 0.6425,
+      "step": 2881
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.559593603945744,
+      "learning_rate": 1.6582718190762718e-05,
+      "loss": 0.7431,
+      "step": 2882
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6447593020281035,
+      "learning_rate": 1.658023783239985e-05,
+      "loss": 0.7389,
+      "step": 2883
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7274509720157214,
+      "learning_rate": 1.657775675985507e-05,
+      "loss": 0.7268,
+      "step": 2884
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6186783596171195,
+      "learning_rate": 1.657527497339766e-05,
+      "loss": 0.8189,
+      "step": 2885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6366378590676298,
+      "learning_rate": 1.6572792473296972e-05,
+      "loss": 0.7788,
+      "step": 2886
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7420947209611584,
+      "learning_rate": 1.6570309259822455e-05,
+      "loss": 0.7549,
+      "step": 2887
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.723075188665114,
+      "learning_rate": 1.656782533324361e-05,
+      "loss": 0.8223,
+      "step": 2888
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6003846657928615,
+      "learning_rate": 1.6565340693830035e-05,
+      "loss": 0.685,
+      "step": 2889
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.509000318762168,
+      "learning_rate": 1.6562855341851402e-05,
+      "loss": 0.6158,
+      "step": 2890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.531501881671066,
+      "learning_rate": 1.6560369277577454e-05,
+      "loss": 0.7816,
+      "step": 2891
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.574341951527415,
+      "learning_rate": 1.6557882501278012e-05,
+      "loss": 0.8377,
+      "step": 2892
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4853515813642724,
+      "learning_rate": 1.6555395013222978e-05,
+      "loss": 0.7056,
+      "step": 2893
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4195032069063818,
+      "learning_rate": 1.6552906813682324e-05,
+      "loss": 0.6399,
+      "step": 2894
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4676046827308649,
+      "learning_rate": 1.655041790292612e-05,
+      "loss": 0.7103,
+      "step": 2895
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5535140253853619,
+      "learning_rate": 1.6547928281224484e-05,
+      "loss": 0.6995,
+      "step": 2896
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4573585724428806,
+      "learning_rate": 1.654543794884764e-05,
+      "loss": 0.793,
+      "step": 2897
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.695730920554382,
+      "learning_rate": 1.6542946906065854e-05,
+      "loss": 0.7337,
+      "step": 2898
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5311418599951014,
+      "learning_rate": 1.654045515314951e-05,
+      "loss": 0.8071,
+      "step": 2899
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6182734204706768,
+      "learning_rate": 1.653796269036904e-05,
+      "loss": 0.8456,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6533053637642925,
+      "learning_rate": 1.653546951799496e-05,
+      "loss": 0.8067,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6011248753907634,
+      "learning_rate": 1.6532975636297863e-05,
+      "loss": 0.7606,
+      "step": 2902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7192000277990367,
+      "learning_rate": 1.6530481045548426e-05,
+      "loss": 0.8073,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.543114296940595,
+      "learning_rate": 1.6527985746017405e-05,
+      "loss": 0.7372,
+      "step": 2904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.431147193220211,
+      "learning_rate": 1.652548973797561e-05,
+      "loss": 0.6621,
+      "step": 2905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4846835227313837,
+      "learning_rate": 1.6522993021693955e-05,
+      "loss": 0.6698,
+      "step": 2906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6667734247359371,
+      "learning_rate": 1.6520495597443415e-05,
+      "loss": 0.7361,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.531655652290317,
+      "learning_rate": 1.6517997465495046e-05,
+      "loss": 0.7556,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6498541138511267,
+      "learning_rate": 1.651549862611998e-05,
+      "loss": 0.8484,
+      "step": 2909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4380001397040072,
+      "learning_rate": 1.6512999079589435e-05,
+      "loss": 0.8083,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5522328190326016,
+      "learning_rate": 1.6510498826174685e-05,
+      "loss": 0.6948,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4467565956725026,
+      "learning_rate": 1.6507997866147107e-05,
+      "loss": 0.644,
+      "step": 2912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5220169878385472,
+      "learning_rate": 1.650549619977813e-05,
+      "loss": 0.747,
+      "step": 2913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5977989579510004,
+      "learning_rate": 1.6502993827339274e-05,
+      "loss": 0.7749,
+      "step": 2914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4482793528657738,
+      "learning_rate": 1.6500490749102137e-05,
+      "loss": 0.7466,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.653687165964818,
+      "learning_rate": 1.6497986965338383e-05,
+      "loss": 0.7703,
+      "step": 2916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.681617825976571,
+      "learning_rate": 1.649548247631976e-05,
+      "loss": 0.7186,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6218352088541688,
+      "learning_rate": 1.64929772823181e-05,
+      "loss": 0.7309,
+      "step": 2918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6085825395587305,
+      "learning_rate": 1.649047138360529e-05,
+      "loss": 0.7282,
+      "step": 2919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5622700051177003,
+      "learning_rate": 1.648796478045331e-05,
+      "loss": 0.6535,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.664828775585605,
+      "learning_rate": 1.648545747313421e-05,
+      "loss": 0.7428,
+      "step": 2921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5657215537870217,
+      "learning_rate": 1.648294946192013e-05,
+      "loss": 0.9065,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4834078087685076,
+      "learning_rate": 1.648044074708326e-05,
+      "loss": 0.6748,
+      "step": 2923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6883500257594792,
+      "learning_rate": 1.6477931328895895e-05,
+      "loss": 0.8581,
+      "step": 2924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6501700073000836,
+      "learning_rate": 1.647542120763039e-05,
+      "loss": 0.783,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.629511650923731,
+      "learning_rate": 1.647291038355917e-05,
+      "loss": 0.8634,
+      "step": 2926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5041014480365311,
+      "learning_rate": 1.6470398856954758e-05,
+      "loss": 0.7276,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6518055797582596,
+      "learning_rate": 1.6467886628089734e-05,
+      "loss": 0.7761,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3929239385366425,
+      "learning_rate": 1.6465373697236762e-05,
+      "loss": 0.6874,
+      "step": 2929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5387621505367852,
+      "learning_rate": 1.6462860064668582e-05,
+      "loss": 0.8216,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7292060414509942,
+      "learning_rate": 1.6460345730658015e-05,
+      "loss": 0.7825,
+      "step": 2931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5574284883149774,
+      "learning_rate": 1.6457830695477942e-05,
+      "loss": 0.7583,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7204118790035463,
+      "learning_rate": 1.6455314959401337e-05,
+      "loss": 0.7306,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5051589890511696,
+      "learning_rate": 1.6452798522701244e-05,
+      "loss": 0.6778,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6934088049746245,
+      "learning_rate": 1.6450281385650783e-05,
+      "loss": 0.7215,
+      "step": 2935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5681140824743331,
+      "learning_rate": 1.6447763548523145e-05,
+      "loss": 0.8215,
+      "step": 2936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.410247781162616,
+      "learning_rate": 1.644524501159161e-05,
+      "loss": 0.6756,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6823167629308104,
+      "learning_rate": 1.644272577512952e-05,
+      "loss": 0.7749,
+      "step": 2938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.630703164194887,
+      "learning_rate": 1.64402058394103e-05,
+      "loss": 0.6726,
+      "step": 2939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.446999359948708,
+      "learning_rate": 1.643768520470745e-05,
+      "loss": 0.6298,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6801001999822267,
+      "learning_rate": 1.6435163871294543e-05,
+      "loss": 0.7354,
+      "step": 2941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5506128809974706,
+      "learning_rate": 1.6432641839445234e-05,
+      "loss": 0.7504,
+      "step": 2942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5430153520759022,
+      "learning_rate": 1.6430119109433247e-05,
+      "loss": 0.7373,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.463452113269399,
+      "learning_rate": 1.6427595681532387e-05,
+      "loss": 0.7108,
+      "step": 2944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4432730436194878,
+      "learning_rate": 1.6425071556016533e-05,
+      "loss": 0.8079,
+      "step": 2945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5186131828255842,
+      "learning_rate": 1.642254673315964e-05,
+      "loss": 0.7656,
+      "step": 2946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4850811230291339,
+      "learning_rate": 1.6420021213235736e-05,
+      "loss": 0.7535,
+      "step": 2947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5396786472697646,
+      "learning_rate": 1.6417494996518925e-05,
+      "loss": 0.7964,
+      "step": 2948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.572772703812896,
+      "learning_rate": 1.641496808328339e-05,
+      "loss": 0.7796,
+      "step": 2949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.621635568935317,
+      "learning_rate": 1.641244047380339e-05,
+      "loss": 0.7015,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6564223885148655,
+      "learning_rate": 1.640991216835326e-05,
+      "loss": 0.7155,
+      "step": 2951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5100958840388738,
+      "learning_rate": 1.64073831672074e-05,
+      "loss": 0.785,
+      "step": 2952
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6905292713911557,
+      "learning_rate": 1.6404853470640295e-05,
+      "loss": 0.7858,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4028782823432704,
+      "learning_rate": 1.640232307892651e-05,
+      "loss": 0.7429,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5817976361332047,
+      "learning_rate": 1.6399791992340674e-05,
+      "loss": 0.7154,
+      "step": 2955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5391480351313862,
+      "learning_rate": 1.63972602111575e-05,
+      "loss": 0.7061,
+      "step": 2956
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6468298584133225,
+      "learning_rate": 1.6394727735651772e-05,
+      "loss": 0.8179,
+      "step": 2957
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6514204198522735,
+      "learning_rate": 1.6392194566098352e-05,
+      "loss": 0.7793,
+      "step": 2958
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5621916853219573,
+      "learning_rate": 1.638966070277217e-05,
+      "loss": 0.7411,
+      "step": 2959
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5437305047711751,
+      "learning_rate": 1.6387126145948248e-05,
+      "loss": 0.8206,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5740494144417758,
+      "learning_rate": 1.638459089590166e-05,
+      "loss": 0.6844,
+      "step": 2961
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4071516788183378,
+      "learning_rate": 1.638205495290758e-05,
+      "loss": 0.7201,
+      "step": 2962
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.566424857115309,
+      "learning_rate": 1.6379518317241236e-05,
+      "loss": 0.616,
+      "step": 2963
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6035080522044143,
+      "learning_rate": 1.6376980989177942e-05,
+      "loss": 0.8079,
+      "step": 2964
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6479440300773858,
+      "learning_rate": 1.6374442968993086e-05,
+      "loss": 0.7135,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4576916199655903,
+      "learning_rate": 1.6371904256962135e-05,
+      "loss": 0.675,
+      "step": 2966
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5816551028695465,
+      "learning_rate": 1.636936485336062e-05,
+      "loss": 0.7392,
+      "step": 2967
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6615128946836017,
+      "learning_rate": 1.6366824758464156e-05,
+      "loss": 0.7514,
+      "step": 2968
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6986189045921958,
+      "learning_rate": 1.636428397254843e-05,
+      "loss": 0.7831,
+      "step": 2969
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6285478666931792,
+      "learning_rate": 1.636174249588921e-05,
+      "loss": 0.8359,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6220292651856945,
+      "learning_rate": 1.635920032876232e-05,
+      "loss": 0.8139,
+      "step": 2971
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5218810852144655,
+      "learning_rate": 1.6356657471443686e-05,
+      "loss": 0.7399,
+      "step": 2972
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.60756563674869,
+      "learning_rate": 1.635411392420929e-05,
+      "loss": 0.719,
+      "step": 2973
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5665830216671,
+      "learning_rate": 1.6351569687335195e-05,
+      "loss": 0.7577,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5933734297319644,
+      "learning_rate": 1.6349024761097532e-05,
+      "loss": 0.7912,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5034799372445287,
+      "learning_rate": 1.634647914577253e-05,
+      "loss": 0.7415,
+      "step": 2976
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4852458185738506,
+      "learning_rate": 1.6343932841636455e-05,
+      "loss": 0.6613,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.473874116603972,
+      "learning_rate": 1.634138584896568e-05,
+      "loss": 0.6285,
+      "step": 2978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.438166471562597,
+      "learning_rate": 1.6338838168036643e-05,
+      "loss": 0.6949,
+      "step": 2979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3987278868685178,
+      "learning_rate": 1.633628979912585e-05,
+      "loss": 0.7365,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6983099446864078,
+      "learning_rate": 1.6333740742509886e-05,
+      "loss": 0.6947,
+      "step": 2981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6141816785702128,
+      "learning_rate": 1.6331190998465416e-05,
+      "loss": 0.7373,
+      "step": 2982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5452457901687615,
+      "learning_rate": 1.632864056726917e-05,
+      "loss": 0.6975,
+      "step": 2983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6608140326198042,
+      "learning_rate": 1.6326089449197956e-05,
+      "loss": 0.7476,
+      "step": 2984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4572606282215208,
+      "learning_rate": 1.6323537644528666e-05,
+      "loss": 0.6997,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5716370053847366,
+      "learning_rate": 1.6320985153538255e-05,
+      "loss": 0.7826,
+      "step": 2986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6592991363088199,
+      "learning_rate": 1.6318431976503754e-05,
+      "loss": 0.7428,
+      "step": 2987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8237988750533487,
+      "learning_rate": 1.6315878113702264e-05,
+      "loss": 0.7558,
+      "step": 2988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6716469257192015,
+      "learning_rate": 1.631332356541098e-05,
+      "loss": 0.7054,
+      "step": 2989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.585305425503702,
+      "learning_rate": 1.6310768331907152e-05,
+      "loss": 0.8262,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5738139196841392,
+      "learning_rate": 1.6308212413468112e-05,
+      "loss": 0.7449,
+      "step": 2991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5846646205419024,
+      "learning_rate": 1.6305655810371263e-05,
+      "loss": 0.7675,
+      "step": 2992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5475748322806462,
+      "learning_rate": 1.630309852289408e-05,
+      "loss": 0.7844,
+      "step": 2993
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4521587234662767,
+      "learning_rate": 1.630054055131413e-05,
+      "loss": 0.9161,
+      "step": 2994
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5509347372810725,
+      "learning_rate": 1.629798189590903e-05,
+      "loss": 0.6987,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.508791759708456,
+      "learning_rate": 1.6295422556956482e-05,
+      "loss": 0.7036,
+      "step": 2996
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6009516480774482,
+      "learning_rate": 1.6292862534734265e-05,
+      "loss": 0.7279,
+      "step": 2997
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4819643251473897,
+      "learning_rate": 1.629030182952023e-05,
+      "loss": 0.7282,
+      "step": 2998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.457259734300266,
+      "learning_rate": 1.6287740441592302e-05,
+      "loss": 0.7124,
+      "step": 2999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.503453385065571,
+      "learning_rate": 1.6285178371228477e-05,
+      "loss": 0.7833,
+      "step": 3000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4651584475008543,
+      "learning_rate": 1.628261561870683e-05,
+      "loss": 0.7883,
+      "step": 3001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5518497406157812,
+      "learning_rate": 1.6280052184305505e-05,
+      "loss": 0.7794,
+      "step": 3002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.487130618612337,
+      "learning_rate": 1.627748806830273e-05,
+      "loss": 0.7968,
+      "step": 3003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8566924238421831,
+      "learning_rate": 1.627492327097679e-05,
+      "loss": 0.7767,
+      "step": 3004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5397235409358518,
+      "learning_rate": 1.627235779260606e-05,
+      "loss": 0.8144,
+      "step": 3005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.688856929862697,
+      "learning_rate": 1.6269791633468975e-05,
+      "loss": 0.8169,
+      "step": 3006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4271495517722819,
+      "learning_rate": 1.626722479384406e-05,
+      "loss": 0.8141,
+      "step": 3007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6141490822849023,
+      "learning_rate": 1.62646572740099e-05,
+      "loss": 0.828,
+      "step": 3008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3862109919366226,
+      "learning_rate": 1.6262089074245163e-05,
+      "loss": 0.6934,
+      "step": 3009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7264246506749796,
+      "learning_rate": 1.6259520194828586e-05,
+      "loss": 0.8084,
+      "step": 3010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6831492410996842,
+      "learning_rate": 1.625695063603898e-05,
+      "loss": 0.7705,
+      "step": 3011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7346547899690183,
+      "learning_rate": 1.6254380398155226e-05,
+      "loss": 0.794,
+      "step": 3012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5155030467406767,
+      "learning_rate": 1.6251809481456292e-05,
+      "loss": 0.7796,
+      "step": 3013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5258784519917301,
+      "learning_rate": 1.6249237886221206e-05,
+      "loss": 0.6925,
+      "step": 3014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7016953394241054,
+      "learning_rate": 1.6246665612729074e-05,
+      "loss": 0.8212,
+      "step": 3015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6296541627190022,
+      "learning_rate": 1.6244092661259073e-05,
+      "loss": 0.7483,
+      "step": 3016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.512531790713189,
+      "learning_rate": 1.6241519032090466e-05,
+      "loss": 0.7023,
+      "step": 3017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6388795320351859,
+      "learning_rate": 1.623894472550257e-05,
+      "loss": 0.8662,
+      "step": 3018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5834486433720267,
+      "learning_rate": 1.623636974177479e-05,
+      "loss": 0.9036,
+      "step": 3019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.608536353251349,
+      "learning_rate": 1.62337940811866e-05,
+      "loss": 0.8445,
+      "step": 3020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6247451048664159,
+      "learning_rate": 1.6231217744017557e-05,
+      "loss": 0.7967,
+      "step": 3021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4633740641195936,
+      "learning_rate": 1.6228640730547264e-05,
+      "loss": 0.8128,
+      "step": 3022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7411844443965356,
+      "learning_rate": 1.622606304105543e-05,
+      "loss": 0.7927,
+      "step": 3023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4140442778396463,
+      "learning_rate": 1.6223484675821813e-05,
+      "loss": 0.6619,
+      "step": 3024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6603407739311253,
+      "learning_rate": 1.622090563512626e-05,
+      "loss": 0.8268,
+      "step": 3025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.457021585971081,
+      "learning_rate": 1.6218325919248688e-05,
+      "loss": 0.7776,
+      "step": 3026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5941908538126683,
+      "learning_rate": 1.6215745528469078e-05,
+      "loss": 0.7932,
+      "step": 3027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6131401763391802,
+      "learning_rate": 1.6213164463067495e-05,
+      "loss": 0.7959,
+      "step": 3028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5040078612577223,
+      "learning_rate": 1.621058272332407e-05,
+      "loss": 0.7749,
+      "step": 3029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4877261332506513,
+      "learning_rate": 1.6208000309519017e-05,
+      "loss": 0.7778,
+      "step": 3030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7057119690601854,
+      "learning_rate": 1.6205417221932612e-05,
+      "loss": 0.8156,
+      "step": 3031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5713357933890586,
+      "learning_rate": 1.620283346084521e-05,
+      "loss": 0.7903,
+      "step": 3032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5128948987964903,
+      "learning_rate": 1.6200249026537232e-05,
+      "loss": 0.6547,
+      "step": 3033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5813368080784047,
+      "learning_rate": 1.619766391928919e-05,
+      "loss": 0.8104,
+      "step": 3034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5360058392310687,
+      "learning_rate": 1.6195078139381647e-05,
+      "loss": 0.7894,
+      "step": 3035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.472545203396127,
+      "learning_rate": 1.619249168709525e-05,
+      "loss": 0.6958,
+      "step": 3036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4597438853991118,
+      "learning_rate": 1.6189904562710722e-05,
+      "loss": 0.6809,
+      "step": 3037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6564332088860414,
+      "learning_rate": 1.6187316766508856e-05,
+      "loss": 0.659,
+      "step": 3038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.526292436196685,
+      "learning_rate": 1.618472829877051e-05,
+      "loss": 0.768,
+      "step": 3039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6314143560380923,
+      "learning_rate": 1.6182139159776627e-05,
+      "loss": 0.8162,
+      "step": 3040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.702836373014158,
+      "learning_rate": 1.6179549349808216e-05,
+      "loss": 0.7858,
+      "step": 3041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.328996213321001,
+      "learning_rate": 1.6176958869146358e-05,
+      "loss": 0.7156,
+      "step": 3042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.567602922565799,
+      "learning_rate": 1.6174367718072213e-05,
+      "loss": 0.729,
+      "step": 3043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5843451063905833,
+      "learning_rate": 1.617177589686701e-05,
+      "loss": 0.8005,
+      "step": 3044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6711460527473359,
+      "learning_rate": 1.6169183405812053e-05,
+      "loss": 0.652,
+      "step": 3045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4961663881076657,
+      "learning_rate": 1.6166590245188708e-05,
+      "loss": 0.6315,
+      "step": 3046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.565609963546912,
+      "learning_rate": 1.6163996415278423e-05,
+      "loss": 0.8349,
+      "step": 3047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3998288323033399,
+      "learning_rate": 1.6161401916362723e-05,
+      "loss": 0.6608,
+      "step": 3048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6135613014089765,
+      "learning_rate": 1.6158806748723205e-05,
+      "loss": 0.7993,
+      "step": 3049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5912260549170647,
+      "learning_rate": 1.6156210912641524e-05,
+      "loss": 0.825,
+      "step": 3050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4735378808073134,
+      "learning_rate": 1.6153614408399418e-05,
+      "loss": 0.6708,
+      "step": 3051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6086824751987387,
+      "learning_rate": 1.61510172362787e-05,
+      "loss": 0.7271,
+      "step": 3052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3443880247958508,
+      "learning_rate": 1.6148419396561254e-05,
+      "loss": 0.7419,
+      "step": 3053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5674892031057017,
+      "learning_rate": 1.6145820889529033e-05,
+      "loss": 0.7514,
+      "step": 3054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5110324653313316,
+      "learning_rate": 1.6143221715464067e-05,
+      "loss": 0.7184,
+      "step": 3055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7002384489332842,
+      "learning_rate": 1.6140621874648447e-05,
+      "loss": 0.7133,
+      "step": 3056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.625221892701305,
+      "learning_rate": 1.6138021367364353e-05,
+      "loss": 0.7698,
+      "step": 3057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.614223008622113,
+      "learning_rate": 1.613542019389403e-05,
+      "loss": 0.8406,
+      "step": 3058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6200348606542283,
+      "learning_rate": 1.613281835451979e-05,
+      "loss": 0.8781,
+      "step": 3059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5587315808852187,
+      "learning_rate": 1.6130215849524025e-05,
+      "loss": 0.7167,
+      "step": 3060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7338243906017512,
+      "learning_rate": 1.6127612679189195e-05,
+      "loss": 0.7895,
+      "step": 3061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7108718796736928,
+      "learning_rate": 1.6125008843797835e-05,
+      "loss": 0.7477,
+      "step": 3062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.592167593368202,
+      "learning_rate": 1.6122404343632547e-05,
+      "loss": 0.7709,
+      "step": 3063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4668296830135767,
+      "learning_rate": 1.6119799178976014e-05,
+      "loss": 0.6438,
+      "step": 3064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6321493961782922,
+      "learning_rate": 1.6117193350110982e-05,
+      "loss": 0.7311,
+      "step": 3065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6487251263538025,
+      "learning_rate": 1.6114586857320272e-05,
+      "loss": 0.675,
+      "step": 3066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7908476169567467,
+      "learning_rate": 1.611197970088678e-05,
+      "loss": 0.8056,
+      "step": 3067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3986976644629658,
+      "learning_rate": 1.6109371881093476e-05,
+      "loss": 0.7328,
+      "step": 3068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6928601199095388,
+      "learning_rate": 1.6106763398223394e-05,
+      "loss": 0.7561,
+      "step": 3069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4871254686923177,
+      "learning_rate": 1.610415425255964e-05,
+      "loss": 0.6538,
+      "step": 3070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.506830990770424,
+      "learning_rate": 1.6101544444385406e-05,
+      "loss": 0.7115,
+      "step": 3071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4004733856804321,
+      "learning_rate": 1.6098933973983934e-05,
+      "loss": 0.7236,
+      "step": 3072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5034428955072574,
+      "learning_rate": 1.609632284163856e-05,
+      "loss": 0.7715,
+      "step": 3073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7482891338939817,
+      "learning_rate": 1.6093711047632676e-05,
+      "loss": 0.7237,
+      "step": 3074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4696133304283667,
+      "learning_rate": 1.6091098592249754e-05,
+      "loss": 0.8095,
+      "step": 3075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6176646068560052,
+      "learning_rate": 1.608848547577333e-05,
+      "loss": 0.7347,
+      "step": 3076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.568225549711669,
+      "learning_rate": 1.6085871698487023e-05,
+      "loss": 0.7641,
+      "step": 3077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.508505589717331,
+      "learning_rate": 1.608325726067452e-05,
+      "loss": 0.7548,
+      "step": 3078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.54588442431525,
+      "learning_rate": 1.6080642162619567e-05,
+      "loss": 0.7671,
+      "step": 3079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6584741334511337,
+      "learning_rate": 1.6078026404605998e-05,
+      "loss": 0.8191,
+      "step": 3080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4583145165427174,
+      "learning_rate": 1.6075409986917714e-05,
+      "loss": 0.7521,
+      "step": 3081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.694055870736409,
+      "learning_rate": 1.6072792909838686e-05,
+      "loss": 0.7129,
+      "step": 3082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6988886090864983,
+      "learning_rate": 1.6070175173652954e-05,
+      "loss": 0.8416,
+      "step": 3083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4580745191676912,
+      "learning_rate": 1.6067556778644633e-05,
+      "loss": 0.6665,
+      "step": 3084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7544466094328668,
+      "learning_rate": 1.606493772509791e-05,
+      "loss": 0.7951,
+      "step": 3085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.53367296871484,
+      "learning_rate": 1.6062318013297045e-05,
+      "loss": 0.8168,
+      "step": 3086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5875483622870659,
+      "learning_rate": 1.6059697643526363e-05,
+      "loss": 0.7529,
+      "step": 3087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4879290031335537,
+      "learning_rate": 1.605707661607026e-05,
+      "loss": 0.7073,
+      "step": 3088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5101875171658965,
+      "learning_rate": 1.6054454931213217e-05,
+      "loss": 0.7343,
+      "step": 3089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4990092041474783,
+      "learning_rate": 1.605183258923977e-05,
+      "loss": 0.7026,
+      "step": 3090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7472358291731696,
+      "learning_rate": 1.6049209590434538e-05,
+      "loss": 0.8645,
+      "step": 3091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5695152177694733,
+      "learning_rate": 1.60465859350822e-05,
+      "loss": 0.7457,
+      "step": 3092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4429394717863138,
+      "learning_rate": 1.6043961623467523e-05,
+      "loss": 0.7753,
+      "step": 3093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4815833238007763,
+      "learning_rate": 1.6041336655875324e-05,
+      "loss": 0.7611,
+      "step": 3094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.653173088460423,
+      "learning_rate": 1.6038711032590507e-05,
+      "loss": 0.7596,
+      "step": 3095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.557819086276001,
+      "learning_rate": 1.6036084753898046e-05,
+      "loss": 0.7968,
+      "step": 3096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4276364100852235,
+      "learning_rate": 1.6033457820082975e-05,
+      "loss": 0.6869,
+      "step": 3097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5354139858886717,
+      "learning_rate": 1.6030830231430412e-05,
+      "loss": 0.8235,
+      "step": 3098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6050709658162097,
+      "learning_rate": 1.6028201988225536e-05,
+      "loss": 0.8106,
+      "step": 3099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6729719675167836,
+      "learning_rate": 1.6025573090753608e-05,
+      "loss": 0.7784,
+      "step": 3100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.599429966910005,
+      "learning_rate": 1.6022943539299948e-05,
+      "loss": 0.7485,
+      "step": 3101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4492624752991685,
+      "learning_rate": 1.602031333414996e-05,
+      "loss": 0.6903,
+      "step": 3102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5228689547530554,
+      "learning_rate": 1.6017682475589103e-05,
+      "loss": 0.7714,
+      "step": 3103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4370981140658128,
+      "learning_rate": 1.601505096390292e-05,
+      "loss": 0.6991,
+      "step": 3104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8961969950247142,
+      "learning_rate": 1.601241879937702e-05,
+      "loss": 0.6681,
+      "step": 3105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.554051609425092,
+      "learning_rate": 1.600978598229708e-05,
+      "loss": 0.7787,
+      "step": 3106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.427848239361892,
+      "learning_rate": 1.6007152512948855e-05,
+      "loss": 0.7268,
+      "step": 3107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5025791642754298,
+      "learning_rate": 1.600451839161817e-05,
+      "loss": 0.7459,
+      "step": 3108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6643850750285802,
+      "learning_rate": 1.6001883618590913e-05,
+      "loss": 0.8121,
+      "step": 3109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.418192732050395,
+      "learning_rate": 1.599924819415305e-05,
+      "loss": 0.7477,
+      "step": 3110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4618962142506864,
+      "learning_rate": 1.5996612118590604e-05,
+      "loss": 0.7099,
+      "step": 3111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.526946571378853,
+      "learning_rate": 1.5993975392189697e-05,
+      "loss": 0.7344,
+      "step": 3112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4919376879669402,
+      "learning_rate": 1.5991338015236494e-05,
+      "loss": 0.8031,
+      "step": 3113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6745537838724252,
+      "learning_rate": 1.5988699988017243e-05,
+      "loss": 0.808,
+      "step": 3114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5838066227108833,
+      "learning_rate": 1.598606131081826e-05,
+      "loss": 0.7917,
+      "step": 3115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5775363708969412,
+      "learning_rate": 1.5983421983925937e-05,
+      "loss": 0.6163,
+      "step": 3116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5748608622317408,
+      "learning_rate": 1.598078200762673e-05,
+      "loss": 0.7859,
+      "step": 3117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5423950467094711,
+      "learning_rate": 1.597814138220716e-05,
+      "loss": 0.7963,
+      "step": 3118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.658696812208415,
+      "learning_rate": 1.597550010795383e-05,
+      "loss": 0.7569,
+      "step": 3119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3567834085980208,
+      "learning_rate": 1.5972858185153412e-05,
+      "loss": 0.6929,
+      "step": 3120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6085663273116662,
+      "learning_rate": 1.5970215614092642e-05,
+      "loss": 0.6935,
+      "step": 3121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3961019816420073,
+      "learning_rate": 1.5967572395058334e-05,
+      "loss": 0.7607,
+      "step": 3122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6165689879778342,
+      "learning_rate": 1.5964928528337363e-05,
+      "loss": 0.8189,
+      "step": 3123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.570775241467543,
+      "learning_rate": 1.596228401421668e-05,
+      "loss": 0.7161,
+      "step": 3124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6292724209145448,
+      "learning_rate": 1.5959638852983306e-05,
+      "loss": 0.7867,
+      "step": 3125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5400550233485224,
+      "learning_rate": 1.5956993044924334e-05,
+      "loss": 0.7101,
+      "step": 3126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7324267822730282,
+      "learning_rate": 1.5954346590326927e-05,
+      "loss": 0.8033,
+      "step": 3127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5495496863750844,
+      "learning_rate": 1.595169948947831e-05,
+      "loss": 0.7431,
+      "step": 3128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6223883659624856,
+      "learning_rate": 1.5949051742665788e-05,
+      "loss": 0.795,
+      "step": 3129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4697103408900418,
+      "learning_rate": 1.594640335017673e-05,
+      "loss": 0.703,
+      "step": 3130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7101405485345647,
+      "learning_rate": 1.5943754312298583e-05,
+      "loss": 0.7555,
+      "step": 3131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5690228773213615,
+      "learning_rate": 1.5941104629318856e-05,
+      "loss": 0.7266,
+      "step": 3132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5873101127509042,
+      "learning_rate": 1.5938454301525126e-05,
+      "loss": 0.7022,
+      "step": 3133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.439819522741593,
+      "learning_rate": 1.593580332920505e-05,
+      "loss": 0.6614,
+      "step": 3134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6108536230936394,
+      "learning_rate": 1.593315171264635e-05,
+      "loss": 0.8291,
+      "step": 3135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4550865485826752,
+      "learning_rate": 1.5930499452136816e-05,
+      "loss": 0.6884,
+      "step": 3136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6520085948866068,
+      "learning_rate": 1.592784654796431e-05,
+      "loss": 0.794,
+      "step": 3137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5170627473468863,
+      "learning_rate": 1.5925193000416756e-05,
+      "loss": 0.6584,
+      "step": 3138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7376109545630536,
+      "learning_rate": 1.5922538809782166e-05,
+      "loss": 0.7054,
+      "step": 3139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5238517632758721,
+      "learning_rate": 1.591988397634861e-05,
+      "loss": 0.7448,
+      "step": 3140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7180058889898717,
+      "learning_rate": 1.591722850040422e-05,
+      "loss": 0.6657,
+      "step": 3141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.562494608534868,
+      "learning_rate": 1.591457238223721e-05,
+      "loss": 0.6182,
+      "step": 3142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5083445543382505,
+      "learning_rate": 1.5911915622135864e-05,
+      "loss": 0.7609,
+      "step": 3143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5367934569040897,
+      "learning_rate": 1.590925822038853e-05,
+      "loss": 0.7238,
+      "step": 3144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5691078613734306,
+      "learning_rate": 1.5906600177283628e-05,
+      "loss": 0.8175,
+      "step": 3145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.690059752667436,
+      "learning_rate": 1.590394149310964e-05,
+      "loss": 0.7669,
+      "step": 3146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.701910126920164,
+      "learning_rate": 1.5901282168155136e-05,
+      "loss": 0.6637,
+      "step": 3147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4938188942808153,
+      "learning_rate": 1.5898622202708734e-05,
+      "loss": 0.7786,
+      "step": 3148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4882713244830919,
+      "learning_rate": 1.5895961597059137e-05,
+      "loss": 0.6403,
+      "step": 3149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.652969857758579,
+      "learning_rate": 1.5893300351495115e-05,
+      "loss": 0.8245,
+      "step": 3150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4934941684919685,
+      "learning_rate": 1.5890638466305495e-05,
+      "loss": 0.7107,
+      "step": 3151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5580866347387254,
+      "learning_rate": 1.5887975941779196e-05,
+      "loss": 0.7022,
+      "step": 3152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3971726739844001,
+      "learning_rate": 1.588531277820518e-05,
+      "loss": 0.6722,
+      "step": 3153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5983800866485998,
+      "learning_rate": 1.58826489758725e-05,
+      "loss": 0.7353,
+      "step": 3154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4430310300597617,
+      "learning_rate": 1.587998453507027e-05,
+      "loss": 0.6149,
+      "step": 3155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5549077993086349,
+      "learning_rate": 1.587731945608767e-05,
+      "loss": 0.7428,
+      "step": 3156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6302877656070816,
+      "learning_rate": 1.5874653739213948e-05,
+      "loss": 0.7145,
+      "step": 3157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4814762359067342,
+      "learning_rate": 1.587198738473844e-05,
+      "loss": 0.7209,
+      "step": 3158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5449844910463155,
+      "learning_rate": 1.5869320392950526e-05,
+      "loss": 0.7601,
+      "step": 3159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5479632840600481,
+      "learning_rate": 1.5866652764139667e-05,
+      "loss": 0.6991,
+      "step": 3160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5665482425525807,
+      "learning_rate": 1.58639844985954e-05,
+      "loss": 0.7769,
+      "step": 3161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4987697650410985,
+      "learning_rate": 1.5861315596607315e-05,
+      "loss": 0.788,
+      "step": 3162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4443483167803732,
+      "learning_rate": 1.585864605846508e-05,
+      "loss": 0.7179,
+      "step": 3163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5872236675585665,
+      "learning_rate": 1.585597588445844e-05,
+      "loss": 0.6736,
+      "step": 3164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5611647818715682,
+      "learning_rate": 1.585330507487719e-05,
+      "loss": 0.8164,
+      "step": 3165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.673138476399332,
+      "learning_rate": 1.585063363001121e-05,
+      "loss": 0.7495,
+      "step": 3166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7099893089128935,
+      "learning_rate": 1.584796155015045e-05,
+      "loss": 0.8088,
+      "step": 3167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4623288158796532,
+      "learning_rate": 1.584528883558491e-05,
+      "loss": 0.7705,
+      "step": 3168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.509934012381232,
+      "learning_rate": 1.5842615486604674e-05,
+      "loss": 0.7781,
+      "step": 3169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5273599096580284,
+      "learning_rate": 1.58399415034999e-05,
+      "loss": 0.823,
+      "step": 3170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5482870979077432,
+      "learning_rate": 1.5837266886560802e-05,
+      "loss": 0.774,
+      "step": 3171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4253493680036589,
+      "learning_rate": 1.583459163607767e-05,
+      "loss": 0.6898,
+      "step": 3172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4850118758223274,
+      "learning_rate": 1.5831915752340855e-05,
+      "loss": 0.7483,
+      "step": 3173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7053278015269768,
+      "learning_rate": 1.5829239235640782e-05,
+      "loss": 0.8189,
+      "step": 3174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4857371360723388,
+      "learning_rate": 1.5826562086267956e-05,
+      "loss": 0.7126,
+      "step": 3175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5745791094329897,
+      "learning_rate": 1.5823884304512934e-05,
+      "loss": 0.7374,
+      "step": 3176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.792000476054425,
+      "learning_rate": 1.582120589066634e-05,
+      "loss": 0.7418,
+      "step": 3177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4355059010070181,
+      "learning_rate": 1.581852684501888e-05,
+      "loss": 0.7174,
+      "step": 3178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6095170706882729,
+      "learning_rate": 1.5815847167861327e-05,
+      "loss": 0.866,
+      "step": 3179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.426623584531756,
+      "learning_rate": 1.5813166859484515e-05,
+      "loss": 0.6396,
+      "step": 3180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3991244579834483,
+      "learning_rate": 1.5810485920179344e-05,
+      "loss": 0.6319,
+      "step": 3181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4587604377004648,
+      "learning_rate": 1.5807804350236793e-05,
+      "loss": 0.6378,
+      "step": 3182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.426280154165666,
+      "learning_rate": 1.5805122149947904e-05,
+      "loss": 0.7459,
+      "step": 3183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5217399746727973,
+      "learning_rate": 1.5802439319603786e-05,
+      "loss": 0.753,
+      "step": 3184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4164338393551477,
+      "learning_rate": 1.5799755859495625e-05,
+      "loss": 0.617,
+      "step": 3185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5097480099291565,
+      "learning_rate": 1.579707176991466e-05,
+      "loss": 0.8471,
+      "step": 3186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.72903010776992,
+      "learning_rate": 1.5794387051152208e-05,
+      "loss": 0.768,
+      "step": 3187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6044825868083277,
+      "learning_rate": 1.5791701703499656e-05,
+      "loss": 0.7313,
+      "step": 3188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5420586309894015,
+      "learning_rate": 1.578901572724846e-05,
+      "loss": 0.6848,
+      "step": 3189
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6985781115712704,
+      "learning_rate": 1.5786329122690135e-05,
+      "loss": 0.728,
+      "step": 3190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3593546053483965,
+      "learning_rate": 1.5783641890116273e-05,
+      "loss": 0.5939,
+      "step": 3191
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.444491418044976,
+      "learning_rate": 1.578095402981853e-05,
+      "loss": 0.6696,
+      "step": 3192
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4372557697332757,
+      "learning_rate": 1.577826554208863e-05,
+      "loss": 0.7297,
+      "step": 3193
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6857280272203305,
+      "learning_rate": 1.577557642721837e-05,
+      "loss": 0.7403,
+      "step": 3194
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5915123912985911,
+      "learning_rate": 1.5772886685499605e-05,
+      "loss": 0.767,
+      "step": 3195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5693985857062513,
+      "learning_rate": 1.5770196317224267e-05,
+      "loss": 0.8271,
+      "step": 3196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7371800022626132,
+      "learning_rate": 1.576750532268436e-05,
+      "loss": 0.7539,
+      "step": 3197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4895456395950455,
+      "learning_rate": 1.5764813702171937e-05,
+      "loss": 0.7717,
+      "step": 3198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5575816426789546,
+      "learning_rate": 1.5762121455979144e-05,
+      "loss": 0.7825,
+      "step": 3199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5366001344613465,
+      "learning_rate": 1.5759428584398175e-05,
+      "loss": 0.7271,
+      "step": 3200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4564051361437806,
+      "learning_rate": 1.5756735087721297e-05,
+      "loss": 0.7732,
+      "step": 3201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4784322680239337,
+      "learning_rate": 1.5754040966240856e-05,
+      "loss": 0.6973,
+      "step": 3202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.547436781038291,
+      "learning_rate": 1.575134622024925e-05,
+      "loss": 0.7008,
+      "step": 3203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6936529434866159,
+      "learning_rate": 1.5748650850038952e-05,
+      "loss": 0.8086,
+      "step": 3204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5107765210344077,
+      "learning_rate": 1.57459548559025e-05,
+      "loss": 0.6434,
+      "step": 3205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4124165547616998,
+      "learning_rate": 1.5743258238132506e-05,
+      "loss": 0.7412,
+      "step": 3206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6505896344500248,
+      "learning_rate": 1.574056099702165e-05,
+      "loss": 0.7154,
+      "step": 3207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3868823830634551,
+      "learning_rate": 1.5737863132862667e-05,
+      "loss": 0.6848,
+      "step": 3208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5852957586226968,
+      "learning_rate": 1.5735164645948365e-05,
+      "loss": 0.719,
+      "step": 3209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5387482663631675,
+      "learning_rate": 1.5732465536571635e-05,
+      "loss": 0.7347,
+      "step": 3210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5964984312736819,
+      "learning_rate": 1.5729765805025416e-05,
+      "loss": 0.7555,
+      "step": 3211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.472043261569749,
+      "learning_rate": 1.5727065451602722e-05,
+      "loss": 0.737,
+      "step": 3212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7394590989294285,
+      "learning_rate": 1.5724364476596637e-05,
+      "loss": 0.8139,
+      "step": 3213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4609690623304736,
+      "learning_rate": 1.5721662880300303e-05,
+      "loss": 0.8106,
+      "step": 3214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4984659414493768,
+      "learning_rate": 1.5718960663006942e-05,
+      "loss": 0.8126,
+      "step": 3215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5078764678998093,
+      "learning_rate": 1.5716257825009836e-05,
+      "loss": 0.7072,
+      "step": 3216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6217296271864472,
+      "learning_rate": 1.5713554366602333e-05,
+      "loss": 0.808,
+      "step": 3217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6107267709731687,
+      "learning_rate": 1.571085028807786e-05,
+      "loss": 0.7409,
+      "step": 3218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.496304836221755,
+      "learning_rate": 1.5708145589729887e-05,
+      "loss": 0.7075,
+      "step": 3219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6943857719430757,
+      "learning_rate": 1.5705440271851982e-05,
+      "loss": 0.82,
+      "step": 3220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5916792170173406,
+      "learning_rate": 1.570273433473776e-05,
+      "loss": 0.7467,
+      "step": 3221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5585203351560093,
+      "learning_rate": 1.57000277786809e-05,
+      "loss": 0.7451,
+      "step": 3222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5284408972175045,
+      "learning_rate": 1.569732060397517e-05,
+      "loss": 0.7689,
+      "step": 3223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.593389917270574,
+      "learning_rate": 1.5694612810914378e-05,
+      "loss": 0.7555,
+      "step": 3224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5395338356559685,
+      "learning_rate": 1.5691904399792425e-05,
+      "loss": 0.7338,
+      "step": 3225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6093287327816626,
+      "learning_rate": 1.5689195370903258e-05,
+      "loss": 0.7608,
+      "step": 3226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5416980247313807,
+      "learning_rate": 1.56864857245409e-05,
+      "loss": 0.6709,
+      "step": 3227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6869407406650818,
+      "learning_rate": 1.5683775460999446e-05,
+      "loss": 0.7468,
+      "step": 3228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5516293625736395,
+      "learning_rate": 1.568106458057305e-05,
+      "loss": 0.7228,
+      "step": 3229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5785437584600943,
+      "learning_rate": 1.5678353083555938e-05,
+      "loss": 0.743,
+      "step": 3230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5095253750379387,
+      "learning_rate": 1.5675640970242393e-05,
+      "loss": 0.7638,
+      "step": 3231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6052697059376566,
+      "learning_rate": 1.5672928240926782e-05,
+      "loss": 0.7752,
+      "step": 3232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6343953052917846,
+      "learning_rate": 1.5670214895903522e-05,
+      "loss": 0.7411,
+      "step": 3233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5285162580731912,
+      "learning_rate": 1.5667500935467112e-05,
+      "loss": 0.866,
+      "step": 3234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.698977154811488,
+      "learning_rate": 1.56647863599121e-05,
+      "loss": 0.7947,
+      "step": 3235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4866475904993897,
+      "learning_rate": 1.566207116953312e-05,
+      "loss": 0.718,
+      "step": 3236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7092189587646751,
+      "learning_rate": 1.5659355364624856e-05,
+      "loss": 0.653,
+      "step": 3237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.503865433297925,
+      "learning_rate": 1.565663894548207e-05,
+      "loss": 0.7405,
+      "step": 3238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4975521811870942,
+      "learning_rate": 1.565392191239959e-05,
+      "loss": 0.7793,
+      "step": 3239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.669709188715031,
+      "learning_rate": 1.5651204265672305e-05,
+      "loss": 0.8376,
+      "step": 3240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6396041162136814,
+      "learning_rate": 1.5648486005595167e-05,
+      "loss": 0.6777,
+      "step": 3241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6140187825941346,
+      "learning_rate": 1.5645767132463205e-05,
+      "loss": 0.7828,
+      "step": 3242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.646485113867051,
+      "learning_rate": 1.5643047646571515e-05,
+      "loss": 0.7514,
+      "step": 3243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.451040577376004,
+      "learning_rate": 1.5640327548215245e-05,
+      "loss": 0.7091,
+      "step": 3244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5039928013791208,
+      "learning_rate": 1.5637606837689632e-05,
+      "loss": 0.6726,
+      "step": 3245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6029000330987817,
+      "learning_rate": 1.563488551528995e-05,
+      "loss": 0.8024,
+      "step": 3246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5792729716232257,
+      "learning_rate": 1.563216358131157e-05,
+      "loss": 0.8164,
+      "step": 3247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.570838961503552,
+      "learning_rate": 1.5629441036049914e-05,
+      "loss": 0.7266,
+      "step": 3248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4417182766874617,
+      "learning_rate": 1.5626717879800465e-05,
+      "loss": 0.7966,
+      "step": 3249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5969231950371279,
+      "learning_rate": 1.562399411285878e-05,
+      "loss": 0.753,
+      "step": 3250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5993321650315608,
+      "learning_rate": 1.5621269735520485e-05,
+      "loss": 0.7534,
+      "step": 3251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.721250803274356,
+      "learning_rate": 1.5618544748081264e-05,
+      "loss": 0.8658,
+      "step": 3252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6157898329866196,
+      "learning_rate": 1.561581915083688e-05,
+      "loss": 0.7707,
+      "step": 3253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4024982634493652,
+      "learning_rate": 1.561309294408315e-05,
+      "loss": 0.6503,
+      "step": 3254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6343740262884703,
+      "learning_rate": 1.5610366128115955e-05,
+      "loss": 0.8926,
+      "step": 3255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6057197627756363,
+      "learning_rate": 1.5607638703231252e-05,
+      "loss": 0.7579,
+      "step": 3256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4338536433313815,
+      "learning_rate": 1.5604910669725066e-05,
+      "loss": 0.7252,
+      "step": 3257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.590164273276705,
+      "learning_rate": 1.5602182027893475e-05,
+      "loss": 0.8045,
+      "step": 3258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7041979020864662,
+      "learning_rate": 1.5599452778032634e-05,
+      "loss": 0.689,
+      "step": 3259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4874237126040357,
+      "learning_rate": 1.559672292043876e-05,
+      "loss": 0.7374,
+      "step": 3260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4925645152044493,
+      "learning_rate": 1.5593992455408136e-05,
+      "loss": 0.6849,
+      "step": 3261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5962062522842229,
+      "learning_rate": 1.559126138323711e-05,
+      "loss": 0.8662,
+      "step": 3262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4143071400598322,
+      "learning_rate": 1.55885297042221e-05,
+      "loss": 0.6973,
+      "step": 3263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4838263696618768,
+      "learning_rate": 1.5585797418659584e-05,
+      "loss": 0.6286,
+      "step": 3264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5525624545464243,
+      "learning_rate": 1.558306452684611e-05,
+      "loss": 0.7891,
+      "step": 3265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4785889855698298,
+      "learning_rate": 1.5580331029078294e-05,
+      "loss": 0.7668,
+      "step": 3266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5508249032227592,
+      "learning_rate": 1.5577596925652808e-05,
+      "loss": 0.6957,
+      "step": 3267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3613392374759203,
+      "learning_rate": 1.5574862216866403e-05,
+      "loss": 0.7151,
+      "step": 3268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5786788804197516,
+      "learning_rate": 1.557212690301588e-05,
+      "loss": 0.6864,
+      "step": 3269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.681522707673013,
+      "learning_rate": 1.5569390984398127e-05,
+      "loss": 0.794,
+      "step": 3270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.682966125140333,
+      "learning_rate": 1.5566654461310073e-05,
+      "loss": 0.7273,
+      "step": 3271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5554113078407963,
+      "learning_rate": 1.556391733404873e-05,
+      "loss": 0.7791,
+      "step": 3272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6006383408949836,
+      "learning_rate": 1.5561179602911173e-05,
+      "loss": 0.7161,
+      "step": 3273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5815849883781743,
+      "learning_rate": 1.5558441268194537e-05,
+      "loss": 0.6785,
+      "step": 3274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4786161669319569,
+      "learning_rate": 1.5555702330196024e-05,
+      "loss": 0.6685,
+      "step": 3275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5620239060191075,
+      "learning_rate": 1.5552962789212904e-05,
+      "loss": 0.7152,
+      "step": 3276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4996886244762353,
+      "learning_rate": 1.5550222645542517e-05,
+      "loss": 0.7554,
+      "step": 3277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6078257952546522,
+      "learning_rate": 1.5547481899482252e-05,
+      "loss": 0.7861,
+      "step": 3278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5363310435057185,
+      "learning_rate": 1.554474055132958e-05,
+      "loss": 0.7803,
+      "step": 3279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5009334366953122,
+      "learning_rate": 1.5541998601382035e-05,
+      "loss": 0.7894,
+      "step": 3280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5874154896475607,
+      "learning_rate": 1.5539256049937205e-05,
+      "loss": 0.6396,
+      "step": 3281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5775813348209935,
+      "learning_rate": 1.5536512897292757e-05,
+      "loss": 0.7288,
+      "step": 3282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6513279054317311,
+      "learning_rate": 1.5533769143746416e-05,
+      "loss": 0.7524,
+      "step": 3283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3792290072180375,
+      "learning_rate": 1.5531024789595968e-05,
+      "loss": 0.6658,
+      "step": 3284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4989661405348915,
+      "learning_rate": 1.552827983513928e-05,
+      "loss": 0.7818,
+      "step": 3285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4834997754072567,
+      "learning_rate": 1.552553428067427e-05,
+      "loss": 0.8217,
+      "step": 3286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.584344077792234,
+      "learning_rate": 1.5522788126498916e-05,
+      "loss": 0.8188,
+      "step": 3287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4302097903103494,
+      "learning_rate": 1.5520041372911286e-05,
+      "loss": 0.7642,
+      "step": 3288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6986211356952792,
+      "learning_rate": 1.5517294020209483e-05,
+      "loss": 0.796,
+      "step": 3289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.670903396086338,
+      "learning_rate": 1.5514546068691697e-05,
+      "loss": 0.8377,
+      "step": 3290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6445195419654928,
+      "learning_rate": 1.5511797518656174e-05,
+      "loss": 0.868,
+      "step": 3291
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6439578769044951,
+      "learning_rate": 1.5509048370401224e-05,
+      "loss": 0.7765,
+      "step": 3292
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.608729331314093,
+      "learning_rate": 1.5506298624225225e-05,
+      "loss": 0.7792,
+      "step": 3293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5707601205056936,
+      "learning_rate": 1.550354828042662e-05,
+      "loss": 0.7926,
+      "step": 3294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.493563824960474,
+      "learning_rate": 1.5500797339303913e-05,
+      "loss": 0.7154,
+      "step": 3295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5552815430878002,
+      "learning_rate": 1.549804580115568e-05,
+      "loss": 0.647,
+      "step": 3296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3313103567796118,
+      "learning_rate": 1.5495293666280556e-05,
+      "loss": 0.7092,
+      "step": 3297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8987395905196398,
+      "learning_rate": 1.549254093497724e-05,
+      "loss": 0.8331,
+      "step": 3298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6075348752193146,
+      "learning_rate": 1.5489787607544498e-05,
+      "loss": 0.8554,
+      "step": 3299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6698223611640848,
+      "learning_rate": 1.5487033684281163e-05,
+      "loss": 0.7023,
+      "step": 3300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5546756934383337,
+      "learning_rate": 1.548427916548613e-05,
+      "loss": 0.882,
+      "step": 3301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5788479756963842,
+      "learning_rate": 1.5481524051458356e-05,
+      "loss": 0.6948,
+      "step": 3302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5474807971285485,
+      "learning_rate": 1.5478768342496872e-05,
+      "loss": 0.8409,
+      "step": 3303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6709675283696996,
+      "learning_rate": 1.547601203890076e-05,
+      "loss": 0.7155,
+      "step": 3304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5105573645817394,
+      "learning_rate": 1.5473255140969176e-05,
+      "loss": 0.6846,
+      "step": 3305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5219539065398735,
+      "learning_rate": 1.5470497649001336e-05,
+      "loss": 0.7485,
+      "step": 3306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5137795824059195,
+      "learning_rate": 1.5467739563296528e-05,
+      "loss": 0.8512,
+      "step": 3307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5675663536182731,
+      "learning_rate": 1.54649808841541e-05,
+      "loss": 0.8345,
+      "step": 3308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.583690197417576,
+      "learning_rate": 1.546222161187345e-05,
+      "loss": 0.7969,
+      "step": 3309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4902741507582222,
+      "learning_rate": 1.5459461746754075e-05,
+      "loss": 0.6601,
+      "step": 3310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5923296984248694,
+      "learning_rate": 1.5456701289095496e-05,
+      "loss": 0.7259,
+      "step": 3311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.569763023547656,
+      "learning_rate": 1.5453940239197328e-05,
+      "loss": 0.7625,
+      "step": 3312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.588716506288604,
+      "learning_rate": 1.5451178597359237e-05,
+      "loss": 0.7954,
+      "step": 3313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.458630633504389,
+      "learning_rate": 1.5448416363880956e-05,
+      "loss": 0.7124,
+      "step": 3314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4807992133698338,
+      "learning_rate": 1.5445653539062283e-05,
+      "loss": 0.8119,
+      "step": 3315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6550269578144803,
+      "learning_rate": 1.5442890123203077e-05,
+      "loss": 0.8211,
+      "step": 3316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6472584346267662,
+      "learning_rate": 1.544012611660327e-05,
+      "loss": 0.7719,
+      "step": 3317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6462001997415807,
+      "learning_rate": 1.5437361519562843e-05,
+      "loss": 0.9011,
+      "step": 3318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6298769830925262,
+      "learning_rate": 1.5434596332381853e-05,
+      "loss": 0.8106,
+      "step": 3319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6925339646508504,
+      "learning_rate": 1.543183055536042e-05,
+      "loss": 0.7336,
+      "step": 3320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5500185607465797,
+      "learning_rate": 1.5429064188798727e-05,
+      "loss": 0.6836,
+      "step": 3321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6311566638530315,
+      "learning_rate": 1.5426297232997016e-05,
+      "loss": 0.6382,
+      "step": 3322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.482678790650085,
+      "learning_rate": 1.5423529688255593e-05,
+      "loss": 0.7339,
+      "step": 3323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4653855741666746,
+      "learning_rate": 1.542076155487484e-05,
+      "loss": 0.8448,
+      "step": 3324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.428470162563532,
+      "learning_rate": 1.54179928331552e-05,
+      "loss": 0.7418,
+      "step": 3325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5905801823778656,
+      "learning_rate": 1.5415223523397153e-05,
+      "loss": 0.7248,
+      "step": 3326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7581095189115765,
+      "learning_rate": 1.541245362590128e-05,
+      "loss": 0.7834,
+      "step": 3327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5621095736065527,
+      "learning_rate": 1.5409683140968213e-05,
+      "loss": 0.7208,
+      "step": 3328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7278372548733352,
+      "learning_rate": 1.540691206889864e-05,
+      "loss": 0.6603,
+      "step": 3329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4516693914835002,
+      "learning_rate": 1.540414040999331e-05,
+      "loss": 0.76,
+      "step": 3330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4918138555527303,
+      "learning_rate": 1.5401368164553054e-05,
+      "loss": 0.7142,
+      "step": 3331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.567648550614192,
+      "learning_rate": 1.539859533287875e-05,
+      "loss": 0.6528,
+      "step": 3332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5803270851499127,
+      "learning_rate": 1.5395821915271344e-05,
+      "loss": 0.7476,
+      "step": 3333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5112348972644198,
+      "learning_rate": 1.539304791203186e-05,
+      "loss": 0.7673,
+      "step": 3334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6631107341011613,
+      "learning_rate": 1.5390273323461354e-05,
+      "loss": 0.7598,
+      "step": 3335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.656394575426843,
+      "learning_rate": 1.5387498149860975e-05,
+      "loss": 0.8525,
+      "step": 3336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4690253202750414,
+      "learning_rate": 1.5384722391531922e-05,
+      "loss": 0.6093,
+      "step": 3337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6346574420144158,
+      "learning_rate": 1.5381946048775462e-05,
+      "loss": 0.7501,
+      "step": 3338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.446909212096057,
+      "learning_rate": 1.5379169121892925e-05,
+      "loss": 0.7256,
+      "step": 3339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4950797036537102,
+      "learning_rate": 1.5376391611185703e-05,
+      "loss": 0.6672,
+      "step": 3340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6664378639633073,
+      "learning_rate": 1.5373613516955243e-05,
+      "loss": 0.7275,
+      "step": 3341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5257888571405214,
+      "learning_rate": 1.537083483950307e-05,
+      "loss": 0.7199,
+      "step": 3342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.484698622122737,
+      "learning_rate": 1.5368055579130768e-05,
+      "loss": 0.7321,
+      "step": 3343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6038707286398843,
+      "learning_rate": 1.5365275736139978e-05,
+      "loss": 0.7752,
+      "step": 3344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.484225545623433,
+      "learning_rate": 1.536249531083241e-05,
+      "loss": 0.655,
+      "step": 3345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4082424654428607,
+      "learning_rate": 1.5359714303509838e-05,
+      "loss": 0.7996,
+      "step": 3346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6178201732198476,
+      "learning_rate": 1.5356932714474096e-05,
+      "loss": 0.7231,
+      "step": 3347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5148520011513111,
+      "learning_rate": 1.535415054402708e-05,
+      "loss": 0.7029,
+      "step": 3348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6893975573942526,
+      "learning_rate": 1.535136779247075e-05,
+      "loss": 0.7314,
+      "step": 3349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5313278686857816,
+      "learning_rate": 1.5348584460107132e-05,
+      "loss": 0.8194,
+      "step": 3350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6152196301716772,
+      "learning_rate": 1.5345800547238315e-05,
+      "loss": 0.7202,
+      "step": 3351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3996437024777064,
+      "learning_rate": 1.534301605416645e-05,
+      "loss": 0.7516,
+      "step": 3352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3860359296983389,
+      "learning_rate": 1.5340230981193745e-05,
+      "loss": 0.6924,
+      "step": 3353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5472239021599388,
+      "learning_rate": 1.5337445328622478e-05,
+      "loss": 0.6571,
+      "step": 3354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6599870981410105,
+      "learning_rate": 1.533465909675499e-05,
+      "loss": 0.7952,
+      "step": 3355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5075921361400568,
+      "learning_rate": 1.5331872285893682e-05,
+      "loss": 0.6703,
+      "step": 3356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4520303260348506,
+      "learning_rate": 1.5329084896341017e-05,
+      "loss": 0.6917,
+      "step": 3357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.509243271112511,
+      "learning_rate": 1.5326296928399527e-05,
+      "loss": 0.7536,
+      "step": 3358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4780434353646639,
+      "learning_rate": 1.5323508382371796e-05,
+      "loss": 0.8208,
+      "step": 3359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7156394695200816,
+      "learning_rate": 1.5320719258560482e-05,
+      "loss": 0.6873,
+      "step": 3360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.569804942520179,
+      "learning_rate": 1.53179295572683e-05,
+      "loss": 0.6311,
+      "step": 3361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.43936954024699,
+      "learning_rate": 1.5315139278798027e-05,
+      "loss": 0.7471,
+      "step": 3362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.531157207106405,
+      "learning_rate": 1.5312348423452506e-05,
+      "loss": 0.7707,
+      "step": 3363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6599684182504466,
+      "learning_rate": 1.5309556991534636e-05,
+      "loss": 0.8001,
+      "step": 3364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7921342018549509,
+      "learning_rate": 1.530676498334739e-05,
+      "loss": 0.7317,
+      "step": 3365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6102687290223747,
+      "learning_rate": 1.5303972399193795e-05,
+      "loss": 0.8126,
+      "step": 3366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7942876571579665,
+      "learning_rate": 1.5301179239376936e-05,
+      "loss": 0.8059,
+      "step": 3367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5295088337332308,
+      "learning_rate": 1.5298385504199975e-05,
+      "loss": 0.6891,
+      "step": 3368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5894517073762104,
+      "learning_rate": 1.5295591193966125e-05,
+      "loss": 0.8017,
+      "step": 3369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5526743477678926,
+      "learning_rate": 1.5292796308978662e-05,
+      "loss": 0.6994,
+      "step": 3370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5052162467039785,
+      "learning_rate": 1.529000084954093e-05,
+      "loss": 0.7152,
+      "step": 3371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6780524182232355,
+      "learning_rate": 1.5287204815956334e-05,
+      "loss": 0.7928,
+      "step": 3372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.683032929660594,
+      "learning_rate": 1.5284408208528342e-05,
+      "loss": 0.755,
+      "step": 3373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6660135300486254,
+      "learning_rate": 1.5281611027560473e-05,
+      "loss": 0.697,
+      "step": 3374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5766118490644017,
+      "learning_rate": 1.527881327335632e-05,
+      "loss": 0.6787,
+      "step": 3375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6779578367810388,
+      "learning_rate": 1.5276014946219546e-05,
+      "loss": 0.919,
+      "step": 3376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6525866035261925,
+      "learning_rate": 1.5273216046453847e-05,
+      "loss": 0.8289,
+      "step": 3377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6010039332657084,
+      "learning_rate": 1.5270416574363015e-05,
+      "loss": 0.7289,
+      "step": 3378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7230029616329539,
+      "learning_rate": 1.5267616530250885e-05,
+      "loss": 0.7253,
+      "step": 3379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4597544106926883,
+      "learning_rate": 1.5264815914421355e-05,
+      "loss": 0.6624,
+      "step": 3380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6520878222696886,
+      "learning_rate": 1.526201472717839e-05,
+      "loss": 0.7278,
+      "step": 3381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.455043108180653,
+      "learning_rate": 1.5259212968826013e-05,
+      "loss": 0.7888,
+      "step": 3382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4135719140088696,
+      "learning_rate": 1.5256410639668317e-05,
+      "loss": 0.6792,
+      "step": 3383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5148945144636952,
+      "learning_rate": 1.5253607740009447e-05,
+      "loss": 0.7648,
+      "step": 3384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3845562724566336,
+      "learning_rate": 1.5250804270153614e-05,
+      "loss": 0.7052,
+      "step": 3385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4782533714227502,
+      "learning_rate": 1.5248000230405086e-05,
+      "loss": 0.8368,
+      "step": 3386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6029221711674988,
+      "learning_rate": 1.5245195621068207e-05,
+      "loss": 0.8218,
+      "step": 3387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7594558175947395,
+      "learning_rate": 1.524239044244737e-05,
+      "loss": 0.8015,
+      "step": 3388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5221214843094437,
+      "learning_rate": 1.5239584694847032e-05,
+      "loss": 0.6736,
+      "step": 3389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6120502289852638,
+      "learning_rate": 1.523677837857171e-05,
+      "loss": 0.7201,
+      "step": 3390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6352747168282848,
+      "learning_rate": 1.523397149392599e-05,
+      "loss": 0.7127,
+      "step": 3391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5988918051963217,
+      "learning_rate": 1.5231164041214515e-05,
+      "loss": 0.868,
+      "step": 3392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5813185864787247,
+      "learning_rate": 1.5228356020741991e-05,
+      "loss": 0.867,
+      "step": 3393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5204761871068275,
+      "learning_rate": 1.5225547432813184e-05,
+      "loss": 0.7359,
+      "step": 3394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4075043465821404,
+      "learning_rate": 1.5222738277732918e-05,
+      "loss": 0.7087,
+      "step": 3395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5446258138687168,
+      "learning_rate": 1.5219928555806087e-05,
+      "loss": 0.7042,
+      "step": 3396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4503770486789294,
+      "learning_rate": 1.5217118267337645e-05,
+      "loss": 0.6952,
+      "step": 3397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5964787717962494,
+      "learning_rate": 1.52143074126326e-05,
+      "loss": 0.709,
+      "step": 3398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6755609053141203,
+      "learning_rate": 1.5211495991996029e-05,
+      "loss": 0.798,
+      "step": 3399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7113459909179223,
+      "learning_rate": 1.5208684005733063e-05,
+      "loss": 0.7412,
+      "step": 3400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4563857410337533,
+      "learning_rate": 1.5205871454148909e-05,
+      "loss": 0.716,
+      "step": 3401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6121262428835532,
+      "learning_rate": 1.5203058337548813e-05,
+      "loss": 0.7882,
+      "step": 3402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6361418926139473,
+      "learning_rate": 1.5200244656238106e-05,
+      "loss": 0.7199,
+      "step": 3403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6940134297858211,
+      "learning_rate": 1.519743041052216e-05,
+      "loss": 0.7211,
+      "step": 3404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.51968529644561,
+      "learning_rate": 1.5194615600706426e-05,
+      "loss": 0.7086,
+      "step": 3405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4702030040840666,
+      "learning_rate": 1.51918002270964e-05,
+      "loss": 0.6855,
+      "step": 3406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.454815808229764,
+      "learning_rate": 1.5188984289997653e-05,
+      "loss": 0.6909,
+      "step": 3407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5142798550349588,
+      "learning_rate": 1.5186167789715805e-05,
+      "loss": 0.7542,
+      "step": 3408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3422009964047386,
+      "learning_rate": 1.5183350726556545e-05,
+      "loss": 0.747,
+      "step": 3409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3183729374682407,
+      "learning_rate": 1.5180533100825625e-05,
+      "loss": 0.7534,
+      "step": 3410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6271445029553226,
+      "learning_rate": 1.5177714912828849e-05,
+      "loss": 0.8038,
+      "step": 3411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4796260691101912,
+      "learning_rate": 1.5174896162872093e-05,
+      "loss": 0.748,
+      "step": 3412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4917898706741968,
+      "learning_rate": 1.5172076851261284e-05,
+      "loss": 0.7499,
+      "step": 3413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4965042302274694,
+      "learning_rate": 1.5169256978302414e-05,
+      "loss": 0.7906,
+      "step": 3414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5864985396719788,
+      "learning_rate": 1.5166436544301537e-05,
+      "loss": 0.7249,
+      "step": 3415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.60300863633716,
+      "learning_rate": 1.5163615549564767e-05,
+      "loss": 0.848,
+      "step": 3416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.522315570698071,
+      "learning_rate": 1.5160793994398279e-05,
+      "loss": 0.7615,
+      "step": 3417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.638295121816351,
+      "learning_rate": 1.5157971879108312e-05,
+      "loss": 0.7377,
+      "step": 3418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6293378893887394,
+      "learning_rate": 1.5155149204001154e-05,
+      "loss": 0.8073,
+      "step": 3419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6763894967757471,
+      "learning_rate": 1.5152325969383173e-05,
+      "loss": 0.6717,
+      "step": 3420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6242355549321144,
+      "learning_rate": 1.5149502175560777e-05,
+      "loss": 0.7134,
+      "step": 3421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4804791823585846,
+      "learning_rate": 1.514667782284045e-05,
+      "loss": 0.6685,
+      "step": 3422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3690241377962513,
+      "learning_rate": 1.5143852911528729e-05,
+      "loss": 0.6693,
+      "step": 3423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.44855052107555,
+      "learning_rate": 1.5141027441932217e-05,
+      "loss": 0.7307,
+      "step": 3424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7367703094119369,
+      "learning_rate": 1.5138201414357574e-05,
+      "loss": 0.6859,
+      "step": 3425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.680051274739284,
+      "learning_rate": 1.513537482911152e-05,
+      "loss": 0.8612,
+      "step": 3426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.678594781233372,
+      "learning_rate": 1.5132547686500834e-05,
+      "loss": 0.7895,
+      "step": 3427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5156833085619492,
+      "learning_rate": 1.5129719986832363e-05,
+      "loss": 0.7615,
+      "step": 3428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5973240474554833,
+      "learning_rate": 1.5126891730413005e-05,
+      "loss": 0.7608,
+      "step": 3429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5179787385027927,
+      "learning_rate": 1.5124062917549728e-05,
+      "loss": 0.7311,
+      "step": 3430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4531619208786193,
+      "learning_rate": 1.512123354854955e-05,
+      "loss": 0.7504,
+      "step": 3431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5130580888869969,
+      "learning_rate": 1.511840362371956e-05,
+      "loss": 0.6863,
+      "step": 3432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4289391942985183,
+      "learning_rate": 1.51155731433669e-05,
+      "loss": 0.7857,
+      "step": 3433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5482410360919439,
+      "learning_rate": 1.511274210779877e-05,
+      "loss": 0.7745,
+      "step": 3434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6338364710983555,
+      "learning_rate": 1.5109910517322442e-05,
+      "loss": 0.7405,
+      "step": 3435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5065075264891135,
+      "learning_rate": 1.5107078372245237e-05,
+      "loss": 0.6536,
+      "step": 3436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.561399673765195,
+      "learning_rate": 1.5104245672874538e-05,
+      "loss": 0.7642,
+      "step": 3437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.516983977202264,
+      "learning_rate": 1.5101412419517796e-05,
+      "loss": 0.7723,
+      "step": 3438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5629150934760518,
+      "learning_rate": 1.509857861248251e-05,
+      "loss": 0.7311,
+      "step": 3439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.642574367936479,
+      "learning_rate": 1.509574425207625e-05,
+      "loss": 0.7182,
+      "step": 3440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.57635743355489,
+      "learning_rate": 1.509290933860664e-05,
+      "loss": 0.7569,
+      "step": 3441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.541793303046407,
+      "learning_rate": 1.5090073872381363e-05,
+      "loss": 0.7027,
+      "step": 3442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5103972648127917,
+      "learning_rate": 1.5087237853708171e-05,
+      "loss": 0.7425,
+      "step": 3443
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4817589947354464,
+      "learning_rate": 1.5084401282894864e-05,
+      "loss": 0.7552,
+      "step": 3444
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.542431128929272,
+      "learning_rate": 1.5081564160249307e-05,
+      "loss": 0.8015,
+      "step": 3445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5698521794455202,
+      "learning_rate": 1.507872648607943e-05,
+      "loss": 0.6561,
+      "step": 3446
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4648673126928915,
+      "learning_rate": 1.5075888260693214e-05,
+      "loss": 0.7357,
+      "step": 3447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4953230949375165,
+      "learning_rate": 1.5073049484398707e-05,
+      "loss": 0.7219,
+      "step": 3448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.497409323484993,
+      "learning_rate": 1.5070210157504009e-05,
+      "loss": 0.7111,
+      "step": 3449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4666121459888177,
+      "learning_rate": 1.5067370280317289e-05,
+      "loss": 0.7245,
+      "step": 3450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7367927555042613,
+      "learning_rate": 1.506452985314677e-05,
+      "loss": 0.7632,
+      "step": 3451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4721834836468148,
+      "learning_rate": 1.5061688876300738e-05,
+      "loss": 0.7348,
+      "step": 3452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.584192423035963,
+      "learning_rate": 1.505884735008753e-05,
+      "loss": 0.739,
+      "step": 3453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5366616006137925,
+      "learning_rate": 1.5056005274815557e-05,
+      "loss": 0.7103,
+      "step": 3454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5065621493020562,
+      "learning_rate": 1.5053162650793277e-05,
+      "loss": 0.8539,
+      "step": 3455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5153782193173004,
+      "learning_rate": 1.5050319478329213e-05,
+      "loss": 0.7514,
+      "step": 3456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.37023829246634,
+      "learning_rate": 1.504747575773195e-05,
+      "loss": 0.7087,
+      "step": 3457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5598410742579687,
+      "learning_rate": 1.5044631489310127e-05,
+      "loss": 0.6599,
+      "step": 3458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4924135931050373,
+      "learning_rate": 1.5041786673372445e-05,
+      "loss": 0.7987,
+      "step": 3459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4892305193365434,
+      "learning_rate": 1.5038941310227663e-05,
+      "loss": 0.7199,
+      "step": 3460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4600408002651917,
+      "learning_rate": 1.5036095400184603e-05,
+      "loss": 0.7402,
+      "step": 3461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.489791988488241,
+      "learning_rate": 1.5033248943552141e-05,
+      "loss": 0.6495,
+      "step": 3462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5042633182167775,
+      "learning_rate": 1.5030401940639221e-05,
+      "loss": 0.7515,
+      "step": 3463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.608923764011562,
+      "learning_rate": 1.5027554391754838e-05,
+      "loss": 0.7703,
+      "step": 3464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4888772998343753,
+      "learning_rate": 1.5024706297208042e-05,
+      "loss": 0.6357,
+      "step": 3465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3819111163671756,
+      "learning_rate": 1.5021857657307957e-05,
+      "loss": 0.6705,
+      "step": 3466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5237889483123492,
+      "learning_rate": 1.5019008472363759e-05,
+      "loss": 0.7462,
+      "step": 3467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5342490853257287,
+      "learning_rate": 1.5016158742684677e-05,
+      "loss": 0.6554,
+      "step": 3468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7384068499426892,
+      "learning_rate": 1.501330846858001e-05,
+      "loss": 0.8208,
+      "step": 3469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5652607294497758,
+      "learning_rate": 1.5010457650359107e-05,
+      "loss": 0.7814,
+      "step": 3470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6478352433592374,
+      "learning_rate": 1.5007606288331382e-05,
+      "loss": 0.7972,
+      "step": 3471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.634555272718592,
+      "learning_rate": 1.50047543828063e-05,
+      "loss": 0.6908,
+      "step": 3472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5720221466031237,
+      "learning_rate": 1.5001901934093401e-05,
+      "loss": 0.7597,
+      "step": 3473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5344693193411723,
+      "learning_rate": 1.4999048942502266e-05,
+      "loss": 0.6344,
+      "step": 3474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4777960019305205,
+      "learning_rate": 1.4996195408342544e-05,
+      "loss": 0.693,
+      "step": 3475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7370996677258854,
+      "learning_rate": 1.4993341331923941e-05,
+      "loss": 0.7798,
+      "step": 3476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5785841948824484,
+      "learning_rate": 1.4990486713556225e-05,
+      "loss": 0.7868,
+      "step": 3477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6036574659873122,
+      "learning_rate": 1.498763155354922e-05,
+      "loss": 0.7637,
+      "step": 3478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7302510182696837,
+      "learning_rate": 1.4984775852212807e-05,
+      "loss": 0.7649,
+      "step": 3479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.664703370876807,
+      "learning_rate": 1.4981919609856927e-05,
+      "loss": 0.7293,
+      "step": 3480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.649517852767688,
+      "learning_rate": 1.4979062826791584e-05,
+      "loss": 0.7203,
+      "step": 3481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4861988697765522,
+      "learning_rate": 1.4976205503326833e-05,
+      "loss": 0.7522,
+      "step": 3482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.536728007961462,
+      "learning_rate": 1.4973347639772796e-05,
+      "loss": 0.6858,
+      "step": 3483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3665924982565034,
+      "learning_rate": 1.4970489236439645e-05,
+      "loss": 0.7136,
+      "step": 3484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7127713453643445,
+      "learning_rate": 1.4967630293637618e-05,
+      "loss": 0.7941,
+      "step": 3485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.523515736800205,
+      "learning_rate": 1.4964770811677007e-05,
+      "loss": 0.7964,
+      "step": 3486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.906325251267563,
+      "learning_rate": 1.4961910790868167e-05,
+      "loss": 0.7163,
+      "step": 3487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4949145245196809,
+      "learning_rate": 1.4959050231521503e-05,
+      "loss": 0.7391,
+      "step": 3488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4994096945124373,
+      "learning_rate": 1.4956189133947492e-05,
+      "loss": 0.664,
+      "step": 3489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5604623367643393,
+      "learning_rate": 1.4953327498456652e-05,
+      "loss": 0.7454,
+      "step": 3490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4918625748558811,
+      "learning_rate": 1.4950465325359573e-05,
+      "loss": 0.8073,
+      "step": 3491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4786728621419427,
+      "learning_rate": 1.4947602614966904e-05,
+      "loss": 0.6621,
+      "step": 3492
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5629535108757933,
+      "learning_rate": 1.494473936758934e-05,
+      "loss": 0.8043,
+      "step": 3493
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.591950029518263,
+      "learning_rate": 1.4941875583537643e-05,
+      "loss": 0.705,
+      "step": 3494
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5452538099171806,
+      "learning_rate": 1.4939011263122635e-05,
+      "loss": 0.6806,
+      "step": 3495
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5899349971218648,
+      "learning_rate": 1.4936146406655194e-05,
+      "loss": 0.8044,
+      "step": 3496
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5583774506657113,
+      "learning_rate": 1.4933281014446253e-05,
+      "loss": 0.7003,
+      "step": 3497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6291525500407988,
+      "learning_rate": 1.4930415086806806e-05,
+      "loss": 0.7825,
+      "step": 3498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.845175889444008,
+      "learning_rate": 1.49275486240479e-05,
+      "loss": 0.8185,
+      "step": 3499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5155979978094534,
+      "learning_rate": 1.4924681626480654e-05,
+      "loss": 0.7676,
+      "step": 3500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5870897903048031,
+      "learning_rate": 1.492181409441623e-05,
+      "loss": 0.7441,
+      "step": 3501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.629353091866416,
+      "learning_rate": 1.4918946028165854e-05,
+      "loss": 0.7711,
+      "step": 3502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.418929425121082,
+      "learning_rate": 1.4916077428040812e-05,
+      "loss": 0.6599,
+      "step": 3503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.78318020605609,
+      "learning_rate": 1.491320829435244e-05,
+      "loss": 0.7912,
+      "step": 3504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.582679788025773,
+      "learning_rate": 1.4910338627412145e-05,
+      "loss": 0.7817,
+      "step": 3505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6395184005956496,
+      "learning_rate": 1.4907468427531378e-05,
+      "loss": 0.7486,
+      "step": 3506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.595727446077858,
+      "learning_rate": 1.4904597695021661e-05,
+      "loss": 0.7169,
+      "step": 3507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7410899348133368,
+      "learning_rate": 1.4901726430194559e-05,
+      "loss": 0.697,
+      "step": 3508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6369873661667074,
+      "learning_rate": 1.489885463336171e-05,
+      "loss": 0.7974,
+      "step": 3509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6216919985187985,
+      "learning_rate": 1.4895982304834801e-05,
+      "loss": 0.8268,
+      "step": 3510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5314031214487973,
+      "learning_rate": 1.4893109444925578e-05,
+      "loss": 0.6685,
+      "step": 3511
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6248727782910444,
+      "learning_rate": 1.4890236053945843e-05,
+      "loss": 0.7778,
+      "step": 3512
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4707089643167897,
+      "learning_rate": 1.4887362132207459e-05,
+      "loss": 0.7438,
+      "step": 3513
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7110333572039853,
+      "learning_rate": 1.4884487680022345e-05,
+      "loss": 0.7783,
+      "step": 3514
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5233667532722868,
+      "learning_rate": 1.4881612697702478e-05,
+      "loss": 0.8079,
+      "step": 3515
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6387517929079978,
+      "learning_rate": 1.4878737185559892e-05,
+      "loss": 0.7378,
+      "step": 3516
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7368485789115762,
+      "learning_rate": 1.4875861143906682e-05,
+      "loss": 0.7348,
+      "step": 3517
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6443544631394493,
+      "learning_rate": 1.4872984573054993e-05,
+      "loss": 0.8044,
+      "step": 3518
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6264571228684506,
+      "learning_rate": 1.4870107473317035e-05,
+      "loss": 0.6913,
+      "step": 3519
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4394243588179805,
+      "learning_rate": 1.4867229845005069e-05,
+      "loss": 0.7788,
+      "step": 3520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5243776042754629,
+      "learning_rate": 1.486435168843142e-05,
+      "loss": 0.8644,
+      "step": 3521
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.39511558840125,
+      "learning_rate": 1.4861473003908464e-05,
+      "loss": 0.7615,
+      "step": 3522
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5237359383852438,
+      "learning_rate": 1.4858593791748636e-05,
+      "loss": 0.7109,
+      "step": 3523
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4141768796323457,
+      "learning_rate": 1.4855714052264436e-05,
+      "loss": 0.6682,
+      "step": 3524
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6013691548953137,
+      "learning_rate": 1.4852833785768411e-05,
+      "loss": 0.7491,
+      "step": 3525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6514226529816114,
+      "learning_rate": 1.4849952992573164e-05,
+      "loss": 0.7559,
+      "step": 3526
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6338844754434587,
+      "learning_rate": 1.4847071672991366e-05,
+      "loss": 0.8517,
+      "step": 3527
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5079211705287743,
+      "learning_rate": 1.4844189827335739e-05,
+      "loss": 0.6482,
+      "step": 3528
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3343098488815823,
+      "learning_rate": 1.484130745591906e-05,
+      "loss": 0.6869,
+      "step": 3529
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6681546973582473,
+      "learning_rate": 1.4838424559054167e-05,
+      "loss": 0.7828,
+      "step": 3530
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7164511219007077,
+      "learning_rate": 1.4835541137053952e-05,
+      "loss": 0.7786,
+      "step": 3531
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5107616051889503,
+      "learning_rate": 1.4832657190231366e-05,
+      "loss": 0.7285,
+      "step": 3532
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.397482883918753,
+      "learning_rate": 1.4829772718899417e-05,
+      "loss": 0.6134,
+      "step": 3533
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.71295749159559,
+      "learning_rate": 1.4826887723371168e-05,
+      "loss": 0.7616,
+      "step": 3534
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6666440172844008,
+      "learning_rate": 1.482400220395974e-05,
+      "loss": 0.7533,
+      "step": 3535
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4688735439647813,
+      "learning_rate": 1.4821116160978313e-05,
+      "loss": 0.7034,
+      "step": 3536
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6611288071734818,
+      "learning_rate": 1.481822959474012e-05,
+      "loss": 0.7442,
+      "step": 3537
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.434539771228801,
+      "learning_rate": 1.4815342505558454e-05,
+      "loss": 0.7714,
+      "step": 3538
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.838877994694268,
+      "learning_rate": 1.4812454893746662e-05,
+      "loss": 0.8659,
+      "step": 3539
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6113005589235647,
+      "learning_rate": 1.4809566759618148e-05,
+      "loss": 0.8147,
+      "step": 3540
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5726735797445943,
+      "learning_rate": 1.4806678103486376e-05,
+      "loss": 0.7031,
+      "step": 3541
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4865889652910735,
+      "learning_rate": 1.4803788925664867e-05,
+      "loss": 0.7913,
+      "step": 3542
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.603946966940541,
+      "learning_rate": 1.480089922646719e-05,
+      "loss": 0.8,
+      "step": 3543
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5696449987816368,
+      "learning_rate": 1.4798009006206979e-05,
+      "loss": 0.892,
+      "step": 3544
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.565736236104443,
+      "learning_rate": 1.4795118265197926e-05,
+      "loss": 0.7458,
+      "step": 3545
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5380747057397648,
+      "learning_rate": 1.479222700375377e-05,
+      "loss": 0.8117,
+      "step": 3546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5974693929548376,
+      "learning_rate": 1.4789335222188314e-05,
+      "loss": 0.7076,
+      "step": 3547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6337144633545488,
+      "learning_rate": 1.478644292081542e-05,
+      "loss": 0.6971,
+      "step": 3548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4854328456568768,
+      "learning_rate": 1.4783550099948995e-05,
+      "loss": 0.6628,
+      "step": 3549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3421312597167916,
+      "learning_rate": 1.4780656759903015e-05,
+      "loss": 0.6936,
+      "step": 3550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5433572587003273,
+      "learning_rate": 1.4777762900991506e-05,
+      "loss": 0.6593,
+      "step": 3551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.524569316576266,
+      "learning_rate": 1.4774868523528548e-05,
+      "loss": 0.7604,
+      "step": 3552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4645164651882978,
+      "learning_rate": 1.4771973627828283e-05,
+      "loss": 0.7331,
+      "step": 3553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6152049411395204,
+      "learning_rate": 1.4769078214204904e-05,
+      "loss": 0.6323,
+      "step": 3554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6252699943112658,
+      "learning_rate": 1.4766182282972667e-05,
+      "loss": 0.8333,
+      "step": 3555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.593900253019558,
+      "learning_rate": 1.4763285834445877e-05,
+      "loss": 0.7247,
+      "step": 3556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4427449082602886,
+      "learning_rate": 1.4760388868938902e-05,
+      "loss": 0.7594,
+      "step": 3557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.556240541058109,
+      "learning_rate": 1.4757491386766154e-05,
+      "loss": 0.8283,
+      "step": 3558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5031696053700674,
+      "learning_rate": 1.4754593388242117e-05,
+      "loss": 0.7985,
+      "step": 3559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4533310803694022,
+      "learning_rate": 1.4751694873681324e-05,
+      "loss": 0.7481,
+      "step": 3560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5333323453412875,
+      "learning_rate": 1.4748795843398361e-05,
+      "loss": 0.6943,
+      "step": 3561
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.389091874942304,
+      "learning_rate": 1.4745896297707867e-05,
+      "loss": 0.6686,
+      "step": 3562
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5586342886882658,
+      "learning_rate": 1.4742996236924551e-05,
+      "loss": 0.7302,
+      "step": 3563
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.478618598982181,
+      "learning_rate": 1.4740095661363165e-05,
+      "loss": 0.7922,
+      "step": 3564
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.626860733600897,
+      "learning_rate": 1.4737194571338523e-05,
+      "loss": 0.7162,
+      "step": 3565
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4377768151388717,
+      "learning_rate": 1.4734292967165491e-05,
+      "loss": 0.7658,
+      "step": 3566
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4427609709577702,
+      "learning_rate": 1.473139084915899e-05,
+      "loss": 0.7086,
+      "step": 3567
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.607204509238383,
+      "learning_rate": 1.4728488217634008e-05,
+      "loss": 0.7084,
+      "step": 3568
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3988616333901625,
+      "learning_rate": 1.4725585072905573e-05,
+      "loss": 0.7179,
+      "step": 3569
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5123617215419016,
+      "learning_rate": 1.4722681415288775e-05,
+      "loss": 0.7475,
+      "step": 3570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.497428536624791,
+      "learning_rate": 1.471977724509877e-05,
+      "loss": 0.7024,
+      "step": 3571
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6175948288740258,
+      "learning_rate": 1.471687256265075e-05,
+      "loss": 0.7575,
+      "step": 3572
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5769398718160916,
+      "learning_rate": 1.4713967368259981e-05,
+      "loss": 0.7064,
+      "step": 3573
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5223116988772043,
+      "learning_rate": 1.4711061662241765e-05,
+      "loss": 0.8851,
+      "step": 3574
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6202519356094804,
+      "learning_rate": 1.4708155444911485e-05,
+      "loss": 0.6783,
+      "step": 3575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5682800582853023,
+      "learning_rate": 1.4705248716584556e-05,
+      "loss": 0.8026,
+      "step": 3576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4970671852419102,
+      "learning_rate": 1.4702341477576461e-05,
+      "loss": 0.7771,
+      "step": 3577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5979231371091105,
+      "learning_rate": 1.4699433728202736e-05,
+      "loss": 0.7805,
+      "step": 3578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.58291483435768,
+      "learning_rate": 1.469652546877897e-05,
+      "loss": 0.8282,
+      "step": 3579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6557268300441905,
+      "learning_rate": 1.4693616699620808e-05,
+      "loss": 0.7751,
+      "step": 3580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5562287512074653,
+      "learning_rate": 1.4690707421043956e-05,
+      "loss": 0.6575,
+      "step": 3581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4175640438046877,
+      "learning_rate": 1.4687797633364167e-05,
+      "loss": 0.6543,
+      "step": 3582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6040932810491193,
+      "learning_rate": 1.4684887336897255e-05,
+      "loss": 0.763,
+      "step": 3583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6058731823369146,
+      "learning_rate": 1.4681976531959085e-05,
+      "loss": 0.8224,
+      "step": 3584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8304388759577574,
+      "learning_rate": 1.4679065218865577e-05,
+      "loss": 0.8874,
+      "step": 3585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4640845041467676,
+      "learning_rate": 1.4676153397932718e-05,
+      "loss": 0.6027,
+      "step": 3586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6107944122500681,
+      "learning_rate": 1.4673241069476532e-05,
+      "loss": 0.7631,
+      "step": 3587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.476859364163601,
+      "learning_rate": 1.4670328233813109e-05,
+      "loss": 0.7166,
+      "step": 3588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.62445852773536,
+      "learning_rate": 1.4667414891258592e-05,
+      "loss": 0.726,
+      "step": 3589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.451038371551322,
+      "learning_rate": 1.4664501042129179e-05,
+      "loss": 0.7385,
+      "step": 3590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.567794040163018,
+      "learning_rate": 1.466158668674112e-05,
+      "loss": 0.6489,
+      "step": 3591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3653304786147358,
+      "learning_rate": 1.4658671825410729e-05,
+      "loss": 0.7955,
+      "step": 3592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.524032132670125,
+      "learning_rate": 1.4655756458454364e-05,
+      "loss": 0.7514,
+      "step": 3593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4538196894282756,
+      "learning_rate": 1.4652840586188442e-05,
+      "loss": 0.7715,
+      "step": 3594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.687136784830419,
+      "learning_rate": 1.4649924208929436e-05,
+      "loss": 0.7911,
+      "step": 3595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4995055304016083,
+      "learning_rate": 1.4647007326993876e-05,
+      "loss": 0.8462,
+      "step": 3596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.403670122516953,
+      "learning_rate": 1.4644089940698341e-05,
+      "loss": 0.6554,
+      "step": 3597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4977699227582575,
+      "learning_rate": 1.4641172050359466e-05,
+      "loss": 0.7081,
+      "step": 3598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6177170296718928,
+      "learning_rate": 1.4638253656293948e-05,
+      "loss": 0.7322,
+      "step": 3599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.562436002230096,
+      "learning_rate": 1.463533475881853e-05,
+      "loss": 0.7545,
+      "step": 3600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5961450775953814,
+      "learning_rate": 1.463241535825001e-05,
+      "loss": 0.7949,
+      "step": 3601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.575564927707248,
+      "learning_rate": 1.4629495454905249e-05,
+      "loss": 0.7422,
+      "step": 3602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.604384087743517,
+      "learning_rate": 1.4626575049101148e-05,
+      "loss": 0.6647,
+      "step": 3603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6632558311955634,
+      "learning_rate": 1.4623654141154682e-05,
+      "loss": 0.8267,
+      "step": 3604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5379881536216085,
+      "learning_rate": 1.4620732731382863e-05,
+      "loss": 0.701,
+      "step": 3605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.562272895182808,
+      "learning_rate": 1.4617810820102766e-05,
+      "loss": 0.7102,
+      "step": 3606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6436077048898663,
+      "learning_rate": 1.461488840763152e-05,
+      "loss": 0.8055,
+      "step": 3607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4710283252621488,
+      "learning_rate": 1.4611965494286304e-05,
+      "loss": 0.7057,
+      "step": 3608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6459419220368325,
+      "learning_rate": 1.4609042080384359e-05,
+      "loss": 0.764,
+      "step": 3609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.618306370841053,
+      "learning_rate": 1.4606118166242974e-05,
+      "loss": 0.8026,
+      "step": 3610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7448311208782217,
+      "learning_rate": 1.4603193752179494e-05,
+      "loss": 0.7705,
+      "step": 3611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.538104497557957,
+      "learning_rate": 1.4600268838511314e-05,
+      "loss": 0.8236,
+      "step": 3612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5925145637942848,
+      "learning_rate": 1.4597343425555894e-05,
+      "loss": 0.737,
+      "step": 3613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5889299388290077,
+      "learning_rate": 1.459441751363074e-05,
+      "loss": 0.7804,
+      "step": 3614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6711569124333012,
+      "learning_rate": 1.4591491103053414e-05,
+      "loss": 0.8212,
+      "step": 3615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.669951186642647,
+      "learning_rate": 1.458856419414153e-05,
+      "loss": 0.8123,
+      "step": 3616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5160115842064925,
+      "learning_rate": 1.458563678721276e-05,
+      "loss": 0.6931,
+      "step": 3617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.577245360536689,
+      "learning_rate": 1.4582708882584831e-05,
+      "loss": 0.8215,
+      "step": 3618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.546826739829446,
+      "learning_rate": 1.4579780480575516e-05,
+      "loss": 0.6896,
+      "step": 3619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6012972025578633,
+      "learning_rate": 1.4576851581502652e-05,
+      "loss": 0.7108,
+      "step": 3620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4358221012271974,
+      "learning_rate": 1.4573922185684118e-05,
+      "loss": 0.7926,
+      "step": 3621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5724954684563723,
+      "learning_rate": 1.4570992293437862e-05,
+      "loss": 0.6527,
+      "step": 3622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4796895310716867,
+      "learning_rate": 1.4568061905081874e-05,
+      "loss": 0.7353,
+      "step": 3623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4753205823693467,
+      "learning_rate": 1.4565131020934203e-05,
+      "loss": 0.7481,
+      "step": 3624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4559061357050045,
+      "learning_rate": 1.456219964131295e-05,
+      "loss": 0.8166,
+      "step": 3625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5246544530422408,
+      "learning_rate": 1.4559267766536272e-05,
+      "loss": 0.8174,
+      "step": 3626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3934216510143873,
+      "learning_rate": 1.4556335396922376e-05,
+      "loss": 0.8245,
+      "step": 3627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.622348369929527,
+      "learning_rate": 1.4553402532789525e-05,
+      "loss": 0.7072,
+      "step": 3628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6293277403360602,
+      "learning_rate": 1.4550469174456038e-05,
+      "loss": 0.8638,
+      "step": 3629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.487813163535852,
+      "learning_rate": 1.4547535322240283e-05,
+      "loss": 0.7255,
+      "step": 3630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5659994707847569,
+      "learning_rate": 1.4544600976460684e-05,
+      "loss": 0.8259,
+      "step": 3631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6386233973682773,
+      "learning_rate": 1.4541666137435717e-05,
+      "loss": 0.896,
+      "step": 3632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4945777648735155,
+      "learning_rate": 1.4538730805483918e-05,
+      "loss": 0.6776,
+      "step": 3633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3869665607423542,
+      "learning_rate": 1.4535794980923863e-05,
+      "loss": 0.6188,
+      "step": 3634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5267935588868653,
+      "learning_rate": 1.45328586640742e-05,
+      "loss": 0.7501,
+      "step": 3635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6141746140610456,
+      "learning_rate": 1.452992185525361e-05,
+      "loss": 0.7758,
+      "step": 3636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7863954147640309,
+      "learning_rate": 1.4526984554780847e-05,
+      "loss": 0.8292,
+      "step": 3637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.437857154083615,
+      "learning_rate": 1.4524046762974705e-05,
+      "loss": 0.7079,
+      "step": 3638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4963813486501298,
+      "learning_rate": 1.4521108480154032e-05,
+      "loss": 0.8067,
+      "step": 3639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6212269153066883,
+      "learning_rate": 1.4518169706637736e-05,
+      "loss": 0.8064,
+      "step": 3640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4496354534684153,
+      "learning_rate": 1.4515230442744774e-05,
+      "loss": 0.6204,
+      "step": 3641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6003373215463383,
+      "learning_rate": 1.4512290688794161e-05,
+      "loss": 0.8154,
+      "step": 3642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4260596812680963,
+      "learning_rate": 1.4509350445104955e-05,
+      "loss": 0.6393,
+      "step": 3643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7932749293783161,
+      "learning_rate": 1.4506409711996278e-05,
+      "loss": 0.7551,
+      "step": 3644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5188216802858843,
+      "learning_rate": 1.45034684897873e-05,
+      "loss": 0.7528,
+      "step": 3645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5663984399125837,
+      "learning_rate": 1.4500526778797244e-05,
+      "loss": 0.7476,
+      "step": 3646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6327098199457324,
+      "learning_rate": 1.4497584579345385e-05,
+      "loss": 0.7634,
+      "step": 3647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5901216906836784,
+      "learning_rate": 1.4494641891751053e-05,
+      "loss": 0.873,
+      "step": 3648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3081865288850023,
+      "learning_rate": 1.4491698716333632e-05,
+      "loss": 0.7266,
+      "step": 3649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4269513171508648,
+      "learning_rate": 1.4488755053412562e-05,
+      "loss": 0.6676,
+      "step": 3650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.844540740181958,
+      "learning_rate": 1.4485810903307324e-05,
+      "loss": 0.8679,
+      "step": 3651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6034428084923098,
+      "learning_rate": 1.4482866266337462e-05,
+      "loss": 0.7452,
+      "step": 3652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6166846651844484,
+      "learning_rate": 1.4479921142822573e-05,
+      "loss": 0.6864,
+      "step": 3653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.464754271434382,
+      "learning_rate": 1.44769755330823e-05,
+      "loss": 0.6319,
+      "step": 3654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5621261510649411,
+      "learning_rate": 1.4474029437436347e-05,
+      "loss": 0.7819,
+      "step": 3655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5159227509879454,
+      "learning_rate": 1.4471082856204464e-05,
+      "loss": 0.7048,
+      "step": 3656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.528622911675774,
+      "learning_rate": 1.4468135789706454e-05,
+      "loss": 0.7776,
+      "step": 3657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5907679003343815,
+      "learning_rate": 1.4465188238262183e-05,
+      "loss": 0.7024,
+      "step": 3658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4917979339422152,
+      "learning_rate": 1.4462240202191553e-05,
+      "loss": 0.8161,
+      "step": 3659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.418213101349719,
+      "learning_rate": 1.4459291681814534e-05,
+      "loss": 0.7786,
+      "step": 3660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4640771784646323,
+      "learning_rate": 1.4456342677451135e-05,
+      "loss": 0.5936,
+      "step": 3661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4022523187744633,
+      "learning_rate": 1.4453393189421428e-05,
+      "loss": 0.7231,
+      "step": 3662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7894244277851525,
+      "learning_rate": 1.4450443218045533e-05,
+      "loss": 0.6724,
+      "step": 3663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5693836473898657,
+      "learning_rate": 1.4447492763643628e-05,
+      "loss": 0.6606,
+      "step": 3664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4248358569829045,
+      "learning_rate": 1.4444541826535932e-05,
+      "loss": 0.7743,
+      "step": 3665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6878677574918814,
+      "learning_rate": 1.4441590407042723e-05,
+      "loss": 0.8351,
+      "step": 3666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5392881799461169,
+      "learning_rate": 1.4438638505484335e-05,
+      "loss": 0.6838,
+      "step": 3667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5517144493417998,
+      "learning_rate": 1.4435686122181152e-05,
+      "loss": 0.7951,
+      "step": 3668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4617789852879355,
+      "learning_rate": 1.4432733257453606e-05,
+      "loss": 0.7206,
+      "step": 3669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5475756651496593,
+      "learning_rate": 1.4429779911622185e-05,
+      "loss": 0.6955,
+      "step": 3670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.586135028957735,
+      "learning_rate": 1.4426826085007429e-05,
+      "loss": 0.7057,
+      "step": 3671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4336651729302952,
+      "learning_rate": 1.442387177792993e-05,
+      "loss": 0.683,
+      "step": 3672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.541184065036639,
+      "learning_rate": 1.442091699071033e-05,
+      "loss": 0.7663,
+      "step": 3673
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.623271185521219,
+      "learning_rate": 1.4417961723669325e-05,
+      "loss": 0.7979,
+      "step": 3674
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.678579444230614,
+      "learning_rate": 1.4415005977127666e-05,
+      "loss": 0.8188,
+      "step": 3675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5397455104574125,
+      "learning_rate": 1.4412049751406149e-05,
+      "loss": 0.7839,
+      "step": 3676
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.858473486282819,
+      "learning_rate": 1.4409093046825628e-05,
+      "loss": 0.9055,
+      "step": 3677
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.45726580316041,
+      "learning_rate": 1.4406135863707011e-05,
+      "loss": 0.7296,
+      "step": 3678
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4610251095613365,
+      "learning_rate": 1.4403178202371246e-05,
+      "loss": 0.6918,
+      "step": 3679
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4127601977062236,
+      "learning_rate": 1.4400220063139348e-05,
+      "loss": 0.7537,
+      "step": 3680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.60772054879461,
+      "learning_rate": 1.4397261446332374e-05,
+      "loss": 0.713,
+      "step": 3681
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6570686371270857,
+      "learning_rate": 1.4394302352271432e-05,
+      "loss": 0.7406,
+      "step": 3682
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6827887598500695,
+      "learning_rate": 1.4391342781277694e-05,
+      "loss": 0.7145,
+      "step": 3683
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4378929980810253,
+      "learning_rate": 1.4388382733672366e-05,
+      "loss": 0.7492,
+      "step": 3684
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5246600418398752,
+      "learning_rate": 1.4385422209776718e-05,
+      "loss": 0.7944,
+      "step": 3685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4014568904240385,
+      "learning_rate": 1.4382461209912073e-05,
+      "loss": 0.6436,
+      "step": 3686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4428423629394982,
+      "learning_rate": 1.4379499734399797e-05,
+      "loss": 0.6961,
+      "step": 3687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.623447139815587,
+      "learning_rate": 1.4376537783561312e-05,
+      "loss": 0.775,
+      "step": 3688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6461117377577357,
+      "learning_rate": 1.4373575357718091e-05,
+      "loss": 0.8546,
+      "step": 3689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.570365184009539,
+      "learning_rate": 1.4370612457191661e-05,
+      "loss": 0.7562,
+      "step": 3690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.525833349657673,
+      "learning_rate": 1.4367649082303598e-05,
+      "loss": 0.6829,
+      "step": 3691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.651551129448428,
+      "learning_rate": 1.4364685233375531e-05,
+      "loss": 0.7618,
+      "step": 3692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8057838334810443,
+      "learning_rate": 1.4361720910729133e-05,
+      "loss": 0.7877,
+      "step": 3693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6881366741148849,
+      "learning_rate": 1.4358756114686144e-05,
+      "loss": 0.7429,
+      "step": 3694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5659478374429834,
+      "learning_rate": 1.4355790845568341e-05,
+      "loss": 0.7668,
+      "step": 3695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8136241690524513,
+      "learning_rate": 1.4352825103697559e-05,
+      "loss": 0.8258,
+      "step": 3696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.547061746400421,
+      "learning_rate": 1.4349858889395682e-05,
+      "loss": 0.8155,
+      "step": 3697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6797520121295821,
+      "learning_rate": 1.4346892202984645e-05,
+      "loss": 0.6994,
+      "step": 3698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4798017009668816,
+      "learning_rate": 1.434392504478644e-05,
+      "loss": 0.7518,
+      "step": 3699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8504089558554946,
+      "learning_rate": 1.4340957415123102e-05,
+      "loss": 0.7717,
+      "step": 3700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4628113248243737,
+      "learning_rate": 1.4337989314316722e-05,
+      "loss": 0.7195,
+      "step": 3701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.718890806307345,
+      "learning_rate": 1.4335020742689439e-05,
+      "loss": 0.777,
+      "step": 3702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6301303689478914,
+      "learning_rate": 1.4332051700563448e-05,
+      "loss": 0.6558,
+      "step": 3703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7492602602785114,
+      "learning_rate": 1.432908218826099e-05,
+      "loss": 0.7529,
+      "step": 3704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5101433076297917,
+      "learning_rate": 1.4326112206104359e-05,
+      "loss": 0.833,
+      "step": 3705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.741038527985199,
+      "learning_rate": 1.4323141754415904e-05,
+      "loss": 0.797,
+      "step": 3706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3241571181008938,
+      "learning_rate": 1.4320170833518012e-05,
+      "loss": 0.632,
+      "step": 3707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5259498416902493,
+      "learning_rate": 1.431719944373314e-05,
+      "loss": 0.7218,
+      "step": 3708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6374207928378792,
+      "learning_rate": 1.4314227585383782e-05,
+      "loss": 0.7591,
+      "step": 3709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4740806885171056,
+      "learning_rate": 1.4311255258792487e-05,
+      "loss": 0.8105,
+      "step": 3710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3374152615652346,
+      "learning_rate": 1.430828246428185e-05,
+      "loss": 0.6014,
+      "step": 3711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.535286610220723,
+      "learning_rate": 1.4305309202174531e-05,
+      "loss": 0.6523,
+      "step": 3712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.62827588808955,
+      "learning_rate": 1.4302335472793223e-05,
+      "loss": 0.8442,
+      "step": 3713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6613625625280515,
+      "learning_rate": 1.4299361276460682e-05,
+      "loss": 0.7732,
+      "step": 3714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.37174038097834,
+      "learning_rate": 1.4296386613499707e-05,
+      "loss": 0.6459,
+      "step": 3715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5986286169846708,
+      "learning_rate": 1.429341148423315e-05,
+      "loss": 0.7608,
+      "step": 3716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.488244744833797,
+      "learning_rate": 1.4290435888983925e-05,
+      "loss": 0.7629,
+      "step": 3717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6315928375643658,
+      "learning_rate": 1.4287459828074974e-05,
+      "loss": 0.7264,
+      "step": 3718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5400859331905221,
+      "learning_rate": 1.4284483301829308e-05,
+      "loss": 0.7096,
+      "step": 3719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3762063201954269,
+      "learning_rate": 1.4281506310569982e-05,
+      "loss": 0.6705,
+      "step": 3720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4196649835515989,
+      "learning_rate": 1.4278528854620101e-05,
+      "loss": 0.6865,
+      "step": 3721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6112878945831668,
+      "learning_rate": 1.4275550934302822e-05,
+      "loss": 0.8048,
+      "step": 3722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.42243294129432,
+      "learning_rate": 1.4272572549941353e-05,
+      "loss": 0.7125,
+      "step": 3723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5105908639260996,
+      "learning_rate": 1.4269593701858946e-05,
+      "loss": 0.7069,
+      "step": 3724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5253097055310054,
+      "learning_rate": 1.4266614390378912e-05,
+      "loss": 0.7579,
+      "step": 3725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6273141589271638,
+      "learning_rate": 1.4263634615824611e-05,
+      "loss": 0.688,
+      "step": 3726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.536711430900752,
+      "learning_rate": 1.4260654378519445e-05,
+      "loss": 0.7457,
+      "step": 3727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6207216793297654,
+      "learning_rate": 1.4257673678786878e-05,
+      "loss": 0.7035,
+      "step": 3728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5322103079932716,
+      "learning_rate": 1.4254692516950415e-05,
+      "loss": 0.6968,
+      "step": 3729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.606998978040395,
+      "learning_rate": 1.4251710893333615e-05,
+      "loss": 0.7688,
+      "step": 3730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6404842672847377,
+      "learning_rate": 1.4248728808260086e-05,
+      "loss": 0.7266,
+      "step": 3731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5499251500162907,
+      "learning_rate": 1.4245746262053489e-05,
+      "loss": 0.815,
+      "step": 3732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4590827206827173,
+      "learning_rate": 1.4242763255037528e-05,
+      "loss": 0.7333,
+      "step": 3733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.520366668123699,
+      "learning_rate": 1.4239779787535969e-05,
+      "loss": 0.7919,
+      "step": 3734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.599862226499546,
+      "learning_rate": 1.4236795859872613e-05,
+      "loss": 0.7795,
+      "step": 3735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5905915358036564,
+      "learning_rate": 1.4233811472371326e-05,
+      "loss": 0.6292,
+      "step": 3736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6291003406413689,
+      "learning_rate": 1.4230826625356011e-05,
+      "loss": 0.7277,
+      "step": 3737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4272235179354023,
+      "learning_rate": 1.422784131915063e-05,
+      "loss": 0.6668,
+      "step": 3738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.832901248039077,
+      "learning_rate": 1.422485555407919e-05,
+      "loss": 0.8227,
+      "step": 3739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7708106990296313,
+      "learning_rate": 1.4221869330465747e-05,
+      "loss": 0.7496,
+      "step": 3740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6247255897769424,
+      "learning_rate": 1.4218882648634413e-05,
+      "loss": 0.7707,
+      "step": 3741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5910916475377026,
+      "learning_rate": 1.4215895508909343e-05,
+      "loss": 0.7974,
+      "step": 3742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4224597783707875,
+      "learning_rate": 1.4212907911614744e-05,
+      "loss": 0.6212,
+      "step": 3743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7028252775857555,
+      "learning_rate": 1.4209919857074875e-05,
+      "loss": 0.7943,
+      "step": 3744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6161012343232768,
+      "learning_rate": 1.420693134561404e-05,
+      "loss": 0.7725,
+      "step": 3745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.678927989956609,
+      "learning_rate": 1.4203942377556596e-05,
+      "loss": 0.846,
+      "step": 3746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3767408598905295,
+      "learning_rate": 1.4200952953226949e-05,
+      "loss": 0.7024,
+      "step": 3747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.545315952518722,
+      "learning_rate": 1.4197963072949556e-05,
+      "loss": 0.7767,
+      "step": 3748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6255543421501542,
+      "learning_rate": 1.419497273704892e-05,
+      "loss": 0.8155,
+      "step": 3749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2918856773106884,
+      "learning_rate": 1.4191981945849595e-05,
+      "loss": 0.6786,
+      "step": 3750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5229528852163654,
+      "learning_rate": 1.4188990699676186e-05,
+      "loss": 0.6854,
+      "step": 3751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.62475224522315,
+      "learning_rate": 1.4185998998853343e-05,
+      "loss": 0.7444,
+      "step": 3752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.458221762720607,
+      "learning_rate": 1.4183006843705774e-05,
+      "loss": 0.702,
+      "step": 3753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4584481686773072,
+      "learning_rate": 1.4180014234558224e-05,
+      "loss": 0.7233,
+      "step": 3754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6129813492541507,
+      "learning_rate": 1.4177021171735499e-05,
+      "loss": 0.6746,
+      "step": 3755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.576899186873065,
+      "learning_rate": 1.4174027655562443e-05,
+      "loss": 0.7028,
+      "step": 3756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4883587909747937,
+      "learning_rate": 1.4171033686363962e-05,
+      "loss": 0.795,
+      "step": 3757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5501029021198707,
+      "learning_rate": 1.4168039264465003e-05,
+      "loss": 0.7463,
+      "step": 3758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.589529021865067,
+      "learning_rate": 1.4165044390190563e-05,
+      "loss": 0.7619,
+      "step": 3759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6718136233085048,
+      "learning_rate": 1.4162049063865686e-05,
+      "loss": 0.7743,
+      "step": 3760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6352469921507213,
+      "learning_rate": 1.4159053285815472e-05,
+      "loss": 0.7488,
+      "step": 3761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.728331817314099,
+      "learning_rate": 1.4156057056365064e-05,
+      "loss": 0.6152,
+      "step": 3762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5786144672305662,
+      "learning_rate": 1.4153060375839656e-05,
+      "loss": 0.8472,
+      "step": 3763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5559091282283846,
+      "learning_rate": 1.4150063244564491e-05,
+      "loss": 0.7672,
+      "step": 3764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6212637710342384,
+      "learning_rate": 1.4147065662864859e-05,
+      "loss": 0.765,
+      "step": 3765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5200602689627594,
+      "learning_rate": 1.4144067631066102e-05,
+      "loss": 0.7064,
+      "step": 3766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.534721277780881,
+      "learning_rate": 1.4141069149493612e-05,
+      "loss": 0.691,
+      "step": 3767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.592590452367276,
+      "learning_rate": 1.4138070218472825e-05,
+      "loss": 0.8375,
+      "step": 3768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4903446030462624,
+      "learning_rate": 1.4135070838329227e-05,
+      "loss": 0.649,
+      "step": 3769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6113849496475525,
+      "learning_rate": 1.4132071009388353e-05,
+      "loss": 0.5889,
+      "step": 3770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5290589043034744,
+      "learning_rate": 1.4129070731975791e-05,
+      "loss": 0.7634,
+      "step": 3771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5352165322416986,
+      "learning_rate": 1.4126070006417174e-05,
+      "loss": 0.7416,
+      "step": 3772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5023754903404172,
+      "learning_rate": 1.412306883303818e-05,
+      "loss": 0.689,
+      "step": 3773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.528870029355672,
+      "learning_rate": 1.4120067212164542e-05,
+      "loss": 0.6352,
+      "step": 3774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4173567433584682,
+      "learning_rate": 1.4117065144122038e-05,
+      "loss": 0.649,
+      "step": 3775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5576072638287743,
+      "learning_rate": 1.4114062629236497e-05,
+      "loss": 0.7539,
+      "step": 3776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5399350550920947,
+      "learning_rate": 1.4111059667833797e-05,
+      "loss": 0.7383,
+      "step": 3777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.574872734967755,
+      "learning_rate": 1.4108056260239858e-05,
+      "loss": 0.7656,
+      "step": 3778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5001125430872018,
+      "learning_rate": 1.4105052406780653e-05,
+      "loss": 0.6585,
+      "step": 3779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7144549791102108,
+      "learning_rate": 1.4102048107782206e-05,
+      "loss": 0.7931,
+      "step": 3780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6329338489197691,
+      "learning_rate": 1.4099043363570588e-05,
+      "loss": 0.8194,
+      "step": 3781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.524154261575747,
+      "learning_rate": 1.4096038174471913e-05,
+      "loss": 0.7184,
+      "step": 3782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5219843735215859,
+      "learning_rate": 1.4093032540812347e-05,
+      "loss": 0.8048,
+      "step": 3783
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4340792225579542,
+      "learning_rate": 1.4090026462918109e-05,
+      "loss": 0.7044,
+      "step": 3784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5796502903346024,
+      "learning_rate": 1.408701994111546e-05,
+      "loss": 0.7784,
+      "step": 3785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.665890566670337,
+      "learning_rate": 1.408401297573071e-05,
+      "loss": 0.7158,
+      "step": 3786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4960904548272018,
+      "learning_rate": 1.4081005567090217e-05,
+      "loss": 0.6829,
+      "step": 3787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4380137590155226,
+      "learning_rate": 1.4077997715520389e-05,
+      "loss": 0.6871,
+      "step": 3788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6775201967511228,
+      "learning_rate": 1.4074989421347683e-05,
+      "loss": 0.7099,
+      "step": 3789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4047990814731268,
+      "learning_rate": 1.4071980684898599e-05,
+      "loss": 0.6937,
+      "step": 3790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5564919382081719,
+      "learning_rate": 1.4068971506499693e-05,
+      "loss": 0.6568,
+      "step": 3791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3846321477736974,
+      "learning_rate": 1.4065961886477561e-05,
+      "loss": 0.6756,
+      "step": 3792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.795205437629554,
+      "learning_rate": 1.4062951825158848e-05,
+      "loss": 0.8043,
+      "step": 3793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.496177531214818,
+      "learning_rate": 1.4059941322870255e-05,
+      "loss": 0.665,
+      "step": 3794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6592131448302958,
+      "learning_rate": 1.405693037993852e-05,
+      "loss": 0.7308,
+      "step": 3795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.587738465551338,
+      "learning_rate": 1.4053918996690437e-05,
+      "loss": 0.8211,
+      "step": 3796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3009250251491375,
+      "learning_rate": 1.4050907173452841e-05,
+      "loss": 0.6029,
+      "step": 3797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.596963960134962,
+      "learning_rate": 1.4047894910552624e-05,
+      "loss": 0.6934,
+      "step": 3798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3562784172634255,
+      "learning_rate": 1.4044882208316714e-05,
+      "loss": 0.7474,
+      "step": 3799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6531533546302477,
+      "learning_rate": 1.4041869067072096e-05,
+      "loss": 0.7492,
+      "step": 3800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6285548654829045,
+      "learning_rate": 1.40388554871458e-05,
+      "loss": 0.8076,
+      "step": 3801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.583551870150882,
+      "learning_rate": 1.4035841468864897e-05,
+      "loss": 0.7962,
+      "step": 3802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.47241936234266,
+      "learning_rate": 1.403282701255652e-05,
+      "loss": 0.7323,
+      "step": 3803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.516670769299008,
+      "learning_rate": 1.4029812118547836e-05,
+      "loss": 0.7883,
+      "step": 3804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5406833413888608,
+      "learning_rate": 1.4026796787166066e-05,
+      "loss": 0.6896,
+      "step": 3805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.521719402197105,
+      "learning_rate": 1.4023781018738474e-05,
+      "loss": 0.6659,
+      "step": 3806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4315441440696073,
+      "learning_rate": 1.402076481359238e-05,
+      "loss": 0.6849,
+      "step": 3807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3956933687096298,
+      "learning_rate": 1.4017748172055146e-05,
+      "loss": 0.7567,
+      "step": 3808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.589580251407946,
+      "learning_rate": 1.4014731094454175e-05,
+      "loss": 0.7499,
+      "step": 3809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4449082249852578,
+      "learning_rate": 1.4011713581116929e-05,
+      "loss": 0.7198,
+      "step": 3810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7331598584411358,
+      "learning_rate": 1.4008695632370905e-05,
+      "loss": 0.771,
+      "step": 3811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5165282302359995,
+      "learning_rate": 1.4005677248543664e-05,
+      "loss": 0.6751,
+      "step": 3812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4090930328374909,
+      "learning_rate": 1.4002658429962797e-05,
+      "loss": 0.7594,
+      "step": 3813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4536552389420134,
+      "learning_rate": 1.3999639176955954e-05,
+      "loss": 0.6469,
+      "step": 3814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6771064874091852,
+      "learning_rate": 1.3996619489850822e-05,
+      "loss": 0.7147,
+      "step": 3815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6093967262983275,
+      "learning_rate": 1.3993599368975148e-05,
+      "loss": 0.7282,
+      "step": 3816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6028915140892204,
+      "learning_rate": 1.3990578814656716e-05,
+      "loss": 0.7347,
+      "step": 3817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5816198883110975,
+      "learning_rate": 1.398755782722336e-05,
+      "loss": 0.6194,
+      "step": 3818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5576591775681445,
+      "learning_rate": 1.3984536407002958e-05,
+      "loss": 0.7161,
+      "step": 3819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.528736190571755,
+      "learning_rate": 1.398151455432344e-05,
+      "loss": 0.7431,
+      "step": 3820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5997533389755794,
+      "learning_rate": 1.397849226951278e-05,
+      "loss": 0.6505,
+      "step": 3821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5242466138313542,
+      "learning_rate": 1.3975469552899003e-05,
+      "loss": 0.7331,
+      "step": 3822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.42476330834485,
+      "learning_rate": 1.3972446404810176e-05,
+      "loss": 0.7084,
+      "step": 3823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.553639315764046,
+      "learning_rate": 1.396942282557441e-05,
+      "loss": 0.828,
+      "step": 3824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.709618287320766,
+      "learning_rate": 1.3966398815519874e-05,
+      "loss": 0.8233,
+      "step": 3825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5133739711007284,
+      "learning_rate": 1.3963374374974774e-05,
+      "loss": 0.688,
+      "step": 3826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6350687600513938,
+      "learning_rate": 1.3960349504267367e-05,
+      "loss": 0.7473,
+      "step": 3827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6231874561776136,
+      "learning_rate": 1.3957324203725952e-05,
+      "loss": 0.7765,
+      "step": 3828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4225649096261395,
+      "learning_rate": 1.395429847367888e-05,
+      "loss": 0.7304,
+      "step": 3829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.524648245289213,
+      "learning_rate": 1.3951272314454549e-05,
+      "loss": 0.7341,
+      "step": 3830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.580200458978781,
+      "learning_rate": 1.39482457263814e-05,
+      "loss": 0.8039,
+      "step": 3831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6674960493863458,
+      "learning_rate": 1.394521870978792e-05,
+      "loss": 0.7352,
+      "step": 3832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5538310665859416,
+      "learning_rate": 1.3942191265002642e-05,
+      "loss": 0.6707,
+      "step": 3833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4450896961970803,
+      "learning_rate": 1.3939163392354156e-05,
+      "loss": 0.623,
+      "step": 3834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7013094745437254,
+      "learning_rate": 1.3936135092171084e-05,
+      "loss": 0.7985,
+      "step": 3835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5946591799629912,
+      "learning_rate": 1.3933106364782103e-05,
+      "loss": 0.7847,
+      "step": 3836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6543035325745643,
+      "learning_rate": 1.393007721051593e-05,
+      "loss": 0.839,
+      "step": 3837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3017152342489924,
+      "learning_rate": 1.3927047629701336e-05,
+      "loss": 0.7176,
+      "step": 3838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4315199274203088,
+      "learning_rate": 1.3924017622667134e-05,
+      "loss": 0.7293,
+      "step": 3839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6881075867618345,
+      "learning_rate": 1.3920987189742186e-05,
+      "loss": 0.7541,
+      "step": 3840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5138340574527112,
+      "learning_rate": 1.3917956331255393e-05,
+      "loss": 0.7273,
+      "step": 3841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6414630893059443,
+      "learning_rate": 1.3914925047535712e-05,
+      "loss": 0.7024,
+      "step": 3842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5294447073641448,
+      "learning_rate": 1.3911893338912142e-05,
+      "loss": 0.7604,
+      "step": 3843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.510644801294499,
+      "learning_rate": 1.390886120571372e-05,
+      "loss": 0.7265,
+      "step": 3844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4877613287490439,
+      "learning_rate": 1.3905828648269544e-05,
+      "loss": 0.6287,
+      "step": 3845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5412889660007723,
+      "learning_rate": 1.390279566690875e-05,
+      "loss": 0.7313,
+      "step": 3846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6011530524353756,
+      "learning_rate": 1.3899762261960519e-05,
+      "loss": 0.7371,
+      "step": 3847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.530472811667168,
+      "learning_rate": 1.3896728433754078e-05,
+      "loss": 0.6527,
+      "step": 3848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4986886699284894,
+      "learning_rate": 1.3893694182618705e-05,
+      "loss": 0.8456,
+      "step": 3849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5593222557486242,
+      "learning_rate": 1.3890659508883719e-05,
+      "loss": 0.7395,
+      "step": 3850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5585342555525534,
+      "learning_rate": 1.3887624412878485e-05,
+      "loss": 0.7369,
+      "step": 3851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6438767824159255,
+      "learning_rate": 1.3884588894932418e-05,
+      "loss": 0.6615,
+      "step": 3852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5879575751740531,
+      "learning_rate": 1.3881552955374975e-05,
+      "loss": 0.7852,
+      "step": 3853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5550148776799544,
+      "learning_rate": 1.3878516594535661e-05,
+      "loss": 0.739,
+      "step": 3854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5185103880024398,
+      "learning_rate": 1.3875479812744022e-05,
+      "loss": 0.803,
+      "step": 3855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.532457978963268,
+      "learning_rate": 1.3872442610329652e-05,
+      "loss": 0.7316,
+      "step": 3856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6130164917900631,
+      "learning_rate": 1.38694049876222e-05,
+      "loss": 0.693,
+      "step": 3857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5063801970280357,
+      "learning_rate": 1.3866366944951344e-05,
+      "loss": 0.7135,
+      "step": 3858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5886342956960429,
+      "learning_rate": 1.386332848264682e-05,
+      "loss": 0.8027,
+      "step": 3859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7263521408368807,
+      "learning_rate": 1.3860289601038406e-05,
+      "loss": 0.7198,
+      "step": 3860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5607732274716484,
+      "learning_rate": 1.3857250300455922e-05,
+      "loss": 0.776,
+      "step": 3861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6904536538120865,
+      "learning_rate": 1.3854210581229242e-05,
+      "loss": 0.7018,
+      "step": 3862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6716662376200253,
+      "learning_rate": 1.3851170443688274e-05,
+      "loss": 0.7247,
+      "step": 3863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4706608496989764,
+      "learning_rate": 1.384812988816298e-05,
+      "loss": 0.6567,
+      "step": 3864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6684571671032502,
+      "learning_rate": 1.3845088914983365e-05,
+      "loss": 0.7876,
+      "step": 3865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5257979153878254,
+      "learning_rate": 1.3842047524479478e-05,
+      "loss": 0.833,
+      "step": 3866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.428668645500322,
+      "learning_rate": 1.3839005716981416e-05,
+      "loss": 0.7895,
+      "step": 3867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.437318096501984,
+      "learning_rate": 1.383596349281932e-05,
+      "loss": 0.7416,
+      "step": 3868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5649894008799563,
+      "learning_rate": 1.383292085232337e-05,
+      "loss": 0.7262,
+      "step": 3869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6585877060908665,
+      "learning_rate": 1.3829877795823805e-05,
+      "loss": 0.679,
+      "step": 3870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6236608561530104,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.6862,
+      "step": 3871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4800539140742295,
+      "learning_rate": 1.3823790436134971e-05,
+      "loss": 0.7337,
+      "step": 3872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4511731411899458,
+      "learning_rate": 1.3820746133606388e-05,
+      "loss": 0.7061,
+      "step": 3873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5955067901139126,
+      "learning_rate": 1.3817701416395562e-05,
+      "loss": 0.7108,
+      "step": 3874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5956323928795686,
+      "learning_rate": 1.381465628483295e-05,
+      "loss": 0.836,
+      "step": 3875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5328985880733839,
+      "learning_rate": 1.3811610739249052e-05,
+      "loss": 0.7159,
+      "step": 3876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.449031123788358,
+      "learning_rate": 1.3808564779974418e-05,
+      "loss": 0.7684,
+      "step": 3877
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.413321842781258,
+      "learning_rate": 1.3805518407339633e-05,
+      "loss": 0.7434,
+      "step": 3878
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6163216631466955,
+      "learning_rate": 1.3802471621675337e-05,
+      "loss": 0.7836,
+      "step": 3879
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5284123049299065,
+      "learning_rate": 1.3799424423312213e-05,
+      "loss": 0.8221,
+      "step": 3880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.459195116401371,
+      "learning_rate": 1.3796376812580983e-05,
+      "loss": 0.6967,
+      "step": 3881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6636941098201041,
+      "learning_rate": 1.379332878981242e-05,
+      "loss": 0.8194,
+      "step": 3882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7712865910914184,
+      "learning_rate": 1.3790280355337332e-05,
+      "loss": 0.7863,
+      "step": 3883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.340039087969083,
+      "learning_rate": 1.378723150948659e-05,
+      "loss": 0.6541,
+      "step": 3884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6035814459079745,
+      "learning_rate": 1.3784182252591091e-05,
+      "loss": 0.8015,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4591240737646074,
+      "learning_rate": 1.3781132584981789e-05,
+      "loss": 0.5822,
+      "step": 3886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4467307457492842,
+      "learning_rate": 1.3778082506989673e-05,
+      "loss": 0.7766,
+      "step": 3887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.495377031381774,
+      "learning_rate": 1.3775032018945784e-05,
+      "loss": 0.6177,
+      "step": 3888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6590640615190357,
+      "learning_rate": 1.3771981121181207e-05,
+      "loss": 0.814,
+      "step": 3889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5529252860548812,
+      "learning_rate": 1.3768929814027064e-05,
+      "loss": 0.7258,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4806719936737924,
+      "learning_rate": 1.376587809781453e-05,
+      "loss": 0.7873,
+      "step": 3891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6248051659767457,
+      "learning_rate": 1.3762825972874816e-05,
+      "loss": 0.6624,
+      "step": 3892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.715735022035157,
+      "learning_rate": 1.375977343953919e-05,
+      "loss": 0.7183,
+      "step": 3893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5299224025682734,
+      "learning_rate": 1.3756720498138956e-05,
+      "loss": 0.7039,
+      "step": 3894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4701266266479829,
+      "learning_rate": 1.3753667149005457e-05,
+      "loss": 0.794,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.567391393980187,
+      "learning_rate": 1.375061339247009e-05,
+      "loss": 0.8436,
+      "step": 3896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6305836041831285,
+      "learning_rate": 1.3747559228864294e-05,
+      "loss": 0.7146,
+      "step": 3897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5033273499462547,
+      "learning_rate": 1.3744504658519545e-05,
+      "loss": 0.707,
+      "step": 3898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.610271975696023,
+      "learning_rate": 1.3741449681767376e-05,
+      "loss": 0.809,
+      "step": 3899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6805097925785062,
+      "learning_rate": 1.3738394298939354e-05,
+      "loss": 0.7389,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5444366150685034,
+      "learning_rate": 1.3735338510367093e-05,
+      "loss": 0.7933,
+      "step": 3901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.488665112098971,
+      "learning_rate": 1.3732282316382249e-05,
+      "loss": 0.7044,
+      "step": 3902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5498518567347663,
+      "learning_rate": 1.3729225717316528e-05,
+      "loss": 0.767,
+      "step": 3903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.465550090058791,
+      "learning_rate": 1.3726168713501673e-05,
+      "loss": 0.7504,
+      "step": 3904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0252785952349286,
+      "learning_rate": 1.3723111305269474e-05,
+      "loss": 0.7166,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7429264453882414,
+      "learning_rate": 1.3720053492951766e-05,
+      "loss": 0.7712,
+      "step": 3906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5456938227408854,
+      "learning_rate": 1.3716995276880428e-05,
+      "loss": 0.6899,
+      "step": 3907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.504687922160811,
+      "learning_rate": 1.3713936657387379e-05,
+      "loss": 0.7122,
+      "step": 3908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5962339911396888,
+      "learning_rate": 1.3710877634804587e-05,
+      "loss": 0.709,
+      "step": 3909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4784206664096584,
+      "learning_rate": 1.3707818209464057e-05,
+      "loss": 0.7649,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.498108568389412,
+      "learning_rate": 1.3704758381697845e-05,
+      "loss": 0.7926,
+      "step": 3911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5681712841089852,
+      "learning_rate": 1.3701698151838048e-05,
+      "loss": 0.6128,
+      "step": 3912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5598006504989548,
+      "learning_rate": 1.3698637520216802e-05,
+      "loss": 0.7507,
+      "step": 3913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4329336123610517,
+      "learning_rate": 1.3695576487166293e-05,
+      "loss": 0.7848,
+      "step": 3914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6529935580307296,
+      "learning_rate": 1.369251505301875e-05,
+      "loss": 0.7724,
+      "step": 3915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4943906668863864,
+      "learning_rate": 1.3689453218106445e-05,
+      "loss": 0.7686,
+      "step": 3916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6263525308946214,
+      "learning_rate": 1.3686390982761689e-05,
+      "loss": 0.7711,
+      "step": 3917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.590883672321078,
+      "learning_rate": 1.368332834731684e-05,
+      "loss": 0.7328,
+      "step": 3918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5011123483847482,
+      "learning_rate": 1.3680265312104297e-05,
+      "loss": 0.6958,
+      "step": 3919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4704528756404118,
+      "learning_rate": 1.367720187745651e-05,
+      "loss": 0.7051,
+      "step": 3920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6128133730605272,
+      "learning_rate": 1.3674138043705962e-05,
+      "loss": 0.9151,
+      "step": 3921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.672605327710779,
+      "learning_rate": 1.367107381118519e-05,
+      "loss": 0.7019,
+      "step": 3922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4984658369655424,
+      "learning_rate": 1.366800918022676e-05,
+      "loss": 0.6323,
+      "step": 3923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.185466425690861,
+      "learning_rate": 1.3664944151163298e-05,
+      "loss": 0.7865,
+      "step": 3924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5925705912270587,
+      "learning_rate": 1.3661878724327462e-05,
+      "loss": 0.7435,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7039222421624558,
+      "learning_rate": 1.3658812900051956e-05,
+      "loss": 0.6535,
+      "step": 3926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5512031904450436,
+      "learning_rate": 1.3655746678669526e-05,
+      "loss": 0.7673,
+      "step": 3927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7229401174639263,
+      "learning_rate": 1.3652680060512965e-05,
+      "loss": 0.7888,
+      "step": 3928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5228373082528688,
+      "learning_rate": 1.3649613045915103e-05,
+      "loss": 0.7251,
+      "step": 3929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6710105183424213,
+      "learning_rate": 1.3646545635208818e-05,
+      "loss": 0.6803,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8352157434527323,
+      "learning_rate": 1.3643477828727032e-05,
+      "loss": 0.7465,
+      "step": 3931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.566225925123017,
+      "learning_rate": 1.3640409626802704e-05,
+      "loss": 0.8061,
+      "step": 3932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.732142431695408,
+      "learning_rate": 1.363734102976884e-05,
+      "loss": 0.8112,
+      "step": 3933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7342905642809994,
+      "learning_rate": 1.3634272037958492e-05,
+      "loss": 0.8466,
+      "step": 3934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.705404804091433,
+      "learning_rate": 1.3631202651704745e-05,
+      "loss": 0.7414,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.481497381352257,
+      "learning_rate": 1.3628132871340738e-05,
+      "loss": 0.7467,
+      "step": 3936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3577307892769122,
+      "learning_rate": 1.3625062697199643e-05,
+      "loss": 0.7035,
+      "step": 3937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.536636098337762,
+      "learning_rate": 1.3621992129614683e-05,
+      "loss": 0.6956,
+      "step": 3938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5967946850115433,
+      "learning_rate": 1.3618921168919119e-05,
+      "loss": 0.7888,
+      "step": 3939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6370373459453844,
+      "learning_rate": 1.3615849815446255e-05,
+      "loss": 0.6748,
+      "step": 3940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4152368310294077,
+      "learning_rate": 1.3612778069529439e-05,
+      "loss": 0.7403,
+      "step": 3941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4061541912594002,
+      "learning_rate": 1.3609705931502063e-05,
+      "loss": 0.6308,
+      "step": 3942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5958246079468728,
+      "learning_rate": 1.3606633401697557e-05,
+      "loss": 0.7026,
+      "step": 3943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5512193775463972,
+      "learning_rate": 1.3603560480449395e-05,
+      "loss": 0.7159,
+      "step": 3944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8355451876396274,
+      "learning_rate": 1.3600487168091099e-05,
+      "loss": 0.8106,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6634340855586849,
+      "learning_rate": 1.3597413464956222e-05,
+      "loss": 0.8114,
+      "step": 3946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5560974007212274,
+      "learning_rate": 1.3594339371378373e-05,
+      "loss": 0.5911,
+      "step": 3947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7429364246328933,
+      "learning_rate": 1.3591264887691198e-05,
+      "loss": 0.7712,
+      "step": 3948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3727463391852266,
+      "learning_rate": 1.3588190014228376e-05,
+      "loss": 0.5769,
+      "step": 3949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5220570359255545,
+      "learning_rate": 1.358511475132364e-05,
+      "loss": 0.7909,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6723979110942109,
+      "learning_rate": 1.3582039099310767e-05,
+      "loss": 0.6782,
+      "step": 3951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.430894102045827,
+      "learning_rate": 1.3578963058523565e-05,
+      "loss": 0.7322,
+      "step": 3952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5004514717420763,
+      "learning_rate": 1.357588662929589e-05,
+      "loss": 0.6998,
+      "step": 3953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4512332645002124,
+      "learning_rate": 1.3572809811961642e-05,
+      "loss": 0.7361,
+      "step": 3954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5840456135063126,
+      "learning_rate": 1.3569732606854764e-05,
+      "loss": 0.7795,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5977379919396377,
+      "learning_rate": 1.3566655014309231e-05,
+      "loss": 0.7411,
+      "step": 3956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6264944495338678,
+      "learning_rate": 1.3563577034659078e-05,
+      "loss": 0.6683,
+      "step": 3957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5164155158232762,
+      "learning_rate": 1.3560498668238364e-05,
+      "loss": 0.7513,
+      "step": 3958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.640008203893221,
+      "learning_rate": 1.35574199153812e-05,
+      "loss": 0.7755,
+      "step": 3959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6039825251834356,
+      "learning_rate": 1.3554340776421736e-05,
+      "loss": 0.6642,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7130422927765,
+      "learning_rate": 1.3551261251694162e-05,
+      "loss": 0.8338,
+      "step": 3961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4141896215299843,
+      "learning_rate": 1.3548181341532715e-05,
+      "loss": 0.7731,
+      "step": 3962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3793364177003844,
+      "learning_rate": 1.3545101046271673e-05,
+      "loss": 0.6984,
+      "step": 3963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.457833073055294,
+      "learning_rate": 1.354202036624535e-05,
+      "loss": 0.7104,
+      "step": 3964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4252227929541943,
+      "learning_rate": 1.353893930178811e-05,
+      "loss": 0.7564,
+      "step": 3965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.476133945396026,
+      "learning_rate": 1.353585785323435e-05,
+      "loss": 0.7568,
+      "step": 3966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5951424524109457,
+      "learning_rate": 1.3532776020918515e-05,
+      "loss": 0.7453,
+      "step": 3967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7376181756282787,
+      "learning_rate": 1.3529693805175087e-05,
+      "loss": 0.7995,
+      "step": 3968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6823680223599915,
+      "learning_rate": 1.3526611206338595e-05,
+      "loss": 0.8328,
+      "step": 3969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5942370022385746,
+      "learning_rate": 1.352352822474361e-05,
+      "loss": 0.6895,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.734311346453895,
+      "learning_rate": 1.3520444860724737e-05,
+      "loss": 0.7768,
+      "step": 3971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4420902721836801,
+      "learning_rate": 1.3517361114616628e-05,
+      "loss": 0.7815,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6730404476991343,
+      "learning_rate": 1.351427698675397e-05,
+      "loss": 0.8037,
+      "step": 3973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7654437159713978,
+      "learning_rate": 1.3511192477471509e-05,
+      "loss": 0.7146,
+      "step": 3974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.645189569094433,
+      "learning_rate": 1.350810758710401e-05,
+      "loss": 0.7678,
+      "step": 3975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5397826998041777,
+      "learning_rate": 1.3505022315986295e-05,
+      "loss": 0.6837,
+      "step": 3976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6013430729945768,
+      "learning_rate": 1.350193666445322e-05,
+      "loss": 0.7314,
+      "step": 3977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.53250292796185,
+      "learning_rate": 1.3498850632839683e-05,
+      "loss": 0.6441,
+      "step": 3978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4275075859196327,
+      "learning_rate": 1.3495764221480625e-05,
+      "loss": 0.6272,
+      "step": 3979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4213206179979412,
+      "learning_rate": 1.349267743071103e-05,
+      "loss": 0.759,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5861061657113282,
+      "learning_rate": 1.3489590260865919e-05,
+      "loss": 0.7647,
+      "step": 3981
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4740057360364553,
+      "learning_rate": 1.3486502712280354e-05,
+      "loss": 0.8151,
+      "step": 3982
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7059409786907052,
+      "learning_rate": 1.3483414785289447e-05,
+      "loss": 0.7177,
+      "step": 3983
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5460171824416302,
+      "learning_rate": 1.3480326480228336e-05,
+      "loss": 0.7102,
+      "step": 3984
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6288133991658376,
+      "learning_rate": 1.3477237797432214e-05,
+      "loss": 0.7615,
+      "step": 3985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5816596029004617,
+      "learning_rate": 1.3474148737236305e-05,
+      "loss": 0.7862,
+      "step": 3986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.472868552210121,
+      "learning_rate": 1.347105929997588e-05,
+      "loss": 0.7079,
+      "step": 3987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5274707854704879,
+      "learning_rate": 1.346796948598625e-05,
+      "loss": 0.749,
+      "step": 3988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6334326446322402,
+      "learning_rate": 1.3464879295602767e-05,
+      "loss": 0.8399,
+      "step": 3989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5472172851604107,
+      "learning_rate": 1.346178872916082e-05,
+      "loss": 0.7571,
+      "step": 3990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4613981141692318,
+      "learning_rate": 1.345869778699584e-05,
+      "loss": 0.6745,
+      "step": 3991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7461178368925054,
+      "learning_rate": 1.3455606469443308e-05,
+      "loss": 0.724,
+      "step": 3992
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5461884023147348,
+      "learning_rate": 1.345251477683873e-05,
+      "loss": 0.7484,
+      "step": 3993
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5143082004134223,
+      "learning_rate": 1.3449422709517665e-05,
+      "loss": 0.6859,
+      "step": 3994
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.472662926327967,
+      "learning_rate": 1.344633026781571e-05,
+      "loss": 0.7049,
+      "step": 3995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.565687265737648,
+      "learning_rate": 1.3443237452068496e-05,
+      "loss": 0.7657,
+      "step": 3996
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4049062205494047,
+      "learning_rate": 1.3440144262611704e-05,
+      "loss": 0.677,
+      "step": 3997
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4465927947849393,
+      "learning_rate": 1.3437050699781052e-05,
+      "loss": 0.6983,
+      "step": 3998
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4138060959994558,
+      "learning_rate": 1.3433956763912293e-05,
+      "loss": 0.6678,
+      "step": 3999
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5221747354708852,
+      "learning_rate": 1.3430862455341228e-05,
+      "loss": 0.7398,
+      "step": 4000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6519117347548133,
+      "learning_rate": 1.3427767774403697e-05,
+      "loss": 0.8253,
+      "step": 4001
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.421518040175894,
+      "learning_rate": 1.3424672721435582e-05,
+      "loss": 0.7104,
+      "step": 4002
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6079084155329708,
+      "learning_rate": 1.3421577296772795e-05,
+      "loss": 0.7406,
+      "step": 4003
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5137819432029258,
+      "learning_rate": 1.34184815007513e-05,
+      "loss": 0.7776,
+      "step": 4004
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7366776196818696,
+      "learning_rate": 1.3415385333707096e-05,
+      "loss": 0.8128,
+      "step": 4005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5077195369673975,
+      "learning_rate": 1.3412288795976228e-05,
+      "loss": 0.705,
+      "step": 4006
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6593668158452635,
+      "learning_rate": 1.340919188789477e-05,
+      "loss": 0.7579,
+      "step": 4007
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5030958680366602,
+      "learning_rate": 1.340609460979885e-05,
+      "loss": 0.7195,
+      "step": 4008
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5556440737593225,
+      "learning_rate": 1.3402996962024623e-05,
+      "loss": 0.7334,
+      "step": 4009
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5110906348055217,
+      "learning_rate": 1.3399898944908292e-05,
+      "loss": 0.7625,
+      "step": 4010
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5609678400026048,
+      "learning_rate": 1.3396800558786102e-05,
+      "loss": 0.6784,
+      "step": 4011
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.687003664548803,
+      "learning_rate": 1.339370180399433e-05,
+      "loss": 0.812,
+      "step": 4012
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5049571514814664,
+      "learning_rate": 1.33906026808693e-05,
+      "loss": 0.7418,
+      "step": 4013
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.639106490446256,
+      "learning_rate": 1.3387503189747369e-05,
+      "loss": 0.6906,
+      "step": 4014
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4993464719464538,
+      "learning_rate": 1.3384403330964944e-05,
+      "loss": 0.7478,
+      "step": 4015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.471817374722621,
+      "learning_rate": 1.3381303104858462e-05,
+      "loss": 0.686,
+      "step": 4016
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.588559439772156,
+      "learning_rate": 1.3378202511764409e-05,
+      "loss": 0.7877,
+      "step": 4017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.623103714420601,
+      "learning_rate": 1.3375101552019296e-05,
+      "loss": 0.7324,
+      "step": 4018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5576566124316074,
+      "learning_rate": 1.3372000225959696e-05,
+      "loss": 0.7979,
+      "step": 4019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6544913152819087,
+      "learning_rate": 1.3368898533922202e-05,
+      "loss": 0.7104,
+      "step": 4020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7688287844335833,
+      "learning_rate": 1.3365796476243455e-05,
+      "loss": 0.731,
+      "step": 4021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.535589463303318,
+      "learning_rate": 1.3362694053260136e-05,
+      "loss": 0.781,
+      "step": 4022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4515536576064387,
+      "learning_rate": 1.3359591265308965e-05,
+      "loss": 0.6553,
+      "step": 4023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6016329573953065,
+      "learning_rate": 1.33564881127267e-05,
+      "loss": 0.7897,
+      "step": 4024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3707604948899,
+      "learning_rate": 1.3353384595850139e-05,
+      "loss": 0.7084,
+      "step": 4025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5991205748896737,
+      "learning_rate": 1.3350280715016122e-05,
+      "loss": 0.7636,
+      "step": 4026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.49161456735389,
+      "learning_rate": 1.3347176470561525e-05,
+      "loss": 0.7854,
+      "step": 4027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.426486035192875,
+      "learning_rate": 1.3344071862823267e-05,
+      "loss": 0.7637,
+      "step": 4028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5661573435509604,
+      "learning_rate": 1.3340966892138306e-05,
+      "loss": 0.6711,
+      "step": 4029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5588267225607866,
+      "learning_rate": 1.3337861558843632e-05,
+      "loss": 0.775,
+      "step": 4030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6100450258007952,
+      "learning_rate": 1.3334755863276288e-05,
+      "loss": 0.7643,
+      "step": 4031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.506099579082188,
+      "learning_rate": 1.3331649805773339e-05,
+      "loss": 0.7668,
+      "step": 4032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6918366583759896,
+      "learning_rate": 1.3328543386671908e-05,
+      "loss": 0.6847,
+      "step": 4033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4958834953303939,
+      "learning_rate": 1.3325436606309146e-05,
+      "loss": 0.6859,
+      "step": 4034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5302344329469577,
+      "learning_rate": 1.3322329465022244e-05,
+      "loss": 0.7167,
+      "step": 4035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4107656193790257,
+      "learning_rate": 1.3319221963148432e-05,
+      "loss": 0.7237,
+      "step": 4036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6684130305180658,
+      "learning_rate": 1.3316114101024984e-05,
+      "loss": 0.7189,
+      "step": 4037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6413025034363493,
+      "learning_rate": 1.331300587898921e-05,
+      "loss": 0.7789,
+      "step": 4038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7096488103304526,
+      "learning_rate": 1.3309897297378456e-05,
+      "loss": 0.8057,
+      "step": 4039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.527070743940658,
+      "learning_rate": 1.3306788356530111e-05,
+      "loss": 0.7914,
+      "step": 4040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.704771614310355,
+      "learning_rate": 1.3303679056781603e-05,
+      "loss": 0.8515,
+      "step": 4041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6883261564971928,
+      "learning_rate": 1.3300569398470398e-05,
+      "loss": 0.8103,
+      "step": 4042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6306291070874364,
+      "learning_rate": 1.3297459381933999e-05,
+      "loss": 0.8424,
+      "step": 4043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6226850730174835,
+      "learning_rate": 1.329434900750995e-05,
+      "loss": 0.6827,
+      "step": 4044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.592943110845101,
+      "learning_rate": 1.3291238275535831e-05,
+      "loss": 0.7318,
+      "step": 4045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6276586615517499,
+      "learning_rate": 1.3288127186349274e-05,
+      "loss": 0.8086,
+      "step": 4046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5809120260098595,
+      "learning_rate": 1.3285015740287926e-05,
+      "loss": 0.6867,
+      "step": 4047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.60571809164613,
+      "learning_rate": 1.3281903937689493e-05,
+      "loss": 0.6884,
+      "step": 4048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4376966388706274,
+      "learning_rate": 1.327879177889171e-05,
+      "loss": 0.6801,
+      "step": 4049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7896708274759923,
+      "learning_rate": 1.3275679264232353e-05,
+      "loss": 0.822,
+      "step": 4050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5150771194204846,
+      "learning_rate": 1.327256639404924e-05,
+      "loss": 0.7674,
+      "step": 4051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5066318786705863,
+      "learning_rate": 1.3269453168680223e-05,
+      "loss": 0.7316,
+      "step": 4052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5224427225579769,
+      "learning_rate": 1.3266339588463191e-05,
+      "loss": 0.7884,
+      "step": 4053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4163623206171054,
+      "learning_rate": 1.3263225653736075e-05,
+      "loss": 0.8494,
+      "step": 4054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5620742435587964,
+      "learning_rate": 1.326011136483685e-05,
+      "loss": 0.7807,
+      "step": 4055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6402310828409679,
+      "learning_rate": 1.3256996722103517e-05,
+      "loss": 0.8428,
+      "step": 4056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4431914952890472,
+      "learning_rate": 1.3253881725874122e-05,
+      "loss": 0.6633,
+      "step": 4057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4191342409961611,
+      "learning_rate": 1.325076637648675e-05,
+      "loss": 0.7194,
+      "step": 4058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6401724364523487,
+      "learning_rate": 1.3247650674279526e-05,
+      "loss": 0.7146,
+      "step": 4059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.955684699842525,
+      "learning_rate": 1.3244534619590608e-05,
+      "loss": 0.7624,
+      "step": 4060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7062570829972066,
+      "learning_rate": 1.3241418212758194e-05,
+      "loss": 0.7219,
+      "step": 4061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3868359029573687,
+      "learning_rate": 1.3238301454120527e-05,
+      "loss": 0.674,
+      "step": 4062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.835307203260098,
+      "learning_rate": 1.3235184344015877e-05,
+      "loss": 0.8416,
+      "step": 4063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9181633969275953,
+      "learning_rate": 1.3232066882782557e-05,
+      "loss": 0.6275,
+      "step": 4064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4805190237294978,
+      "learning_rate": 1.3228949070758925e-05,
+      "loss": 0.6706,
+      "step": 4065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6628355377811277,
+      "learning_rate": 1.3225830908283363e-05,
+      "loss": 0.7272,
+      "step": 4066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.525776708814196,
+      "learning_rate": 1.3222712395694303e-05,
+      "loss": 0.7371,
+      "step": 4067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5025346264582065,
+      "learning_rate": 1.3219593533330211e-05,
+      "loss": 0.8229,
+      "step": 4068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7879124391361025,
+      "learning_rate": 1.3216474321529589e-05,
+      "loss": 0.7747,
+      "step": 4069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6357297829478563,
+      "learning_rate": 1.3213354760630981e-05,
+      "loss": 0.6958,
+      "step": 4070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5215139989575535,
+      "learning_rate": 1.3210234850972966e-05,
+      "loss": 0.7636,
+      "step": 4071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3885791150732956,
+      "learning_rate": 1.3207114592894155e-05,
+      "loss": 0.6559,
+      "step": 4072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5713248008518697,
+      "learning_rate": 1.3203993986733212e-05,
+      "loss": 0.719,
+      "step": 4073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5446063043547251,
+      "learning_rate": 1.3200873032828827e-05,
+      "loss": 0.7433,
+      "step": 4074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.564066324834634,
+      "learning_rate": 1.319775173151973e-05,
+      "loss": 0.6929,
+      "step": 4075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6681039151887862,
+      "learning_rate": 1.319463008314469e-05,
+      "loss": 0.811,
+      "step": 4076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7337500036471722,
+      "learning_rate": 1.3191508088042511e-05,
+      "loss": 0.7997,
+      "step": 4077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5590561891101518,
+      "learning_rate": 1.3188385746552039e-05,
+      "loss": 0.8167,
+      "step": 4078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6649219606880605,
+      "learning_rate": 1.3185263059012158e-05,
+      "loss": 0.7256,
+      "step": 4079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4916284993073423,
+      "learning_rate": 1.3182140025761782e-05,
+      "loss": 0.7404,
+      "step": 4080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5809091122227412,
+      "learning_rate": 1.3179016647139866e-05,
+      "loss": 0.6229,
+      "step": 4081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.577754747739406,
+      "learning_rate": 1.3175892923485411e-05,
+      "loss": 0.7157,
+      "step": 4082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.494271837332332,
+      "learning_rate": 1.3172768855137446e-05,
+      "loss": 0.866,
+      "step": 4083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.591720798787326,
+      "learning_rate": 1.3169644442435036e-05,
+      "loss": 0.7013,
+      "step": 4084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5553585860498986,
+      "learning_rate": 1.316651968571729e-05,
+      "loss": 0.8126,
+      "step": 4085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4891715960594787,
+      "learning_rate": 1.316339458532335e-05,
+      "loss": 0.7398,
+      "step": 4086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.542762990549468,
+      "learning_rate": 1.31602691415924e-05,
+      "loss": 0.7326,
+      "step": 4087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.490050148493316,
+      "learning_rate": 1.3157143354863653e-05,
+      "loss": 0.7312,
+      "step": 4088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7648769404474558,
+      "learning_rate": 1.315401722547637e-05,
+      "loss": 0.7986,
+      "step": 4089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.719709018178479,
+      "learning_rate": 1.3150890753769836e-05,
+      "loss": 0.7715,
+      "step": 4090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5627320403532918,
+      "learning_rate": 1.314776394008339e-05,
+      "loss": 0.7837,
+      "step": 4091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.453463710193572,
+      "learning_rate": 1.3144636784756391e-05,
+      "loss": 0.6246,
+      "step": 4092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.650974176196651,
+      "learning_rate": 1.3141509288128248e-05,
+      "loss": 0.786,
+      "step": 4093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5484033252730638,
+      "learning_rate": 1.3138381450538398e-05,
+      "loss": 0.7412,
+      "step": 4094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6037113882082326,
+      "learning_rate": 1.3135253272326316e-05,
+      "loss": 0.6865,
+      "step": 4095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5918578191023562,
+      "learning_rate": 1.3132124753831525e-05,
+      "loss": 0.8283,
+      "step": 4096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4520708558181288,
+      "learning_rate": 1.3128995895393576e-05,
+      "loss": 0.7158,
+      "step": 4097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4518469912628422,
+      "learning_rate": 1.3125866697352052e-05,
+      "loss": 0.7374,
+      "step": 4098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5362404511239431,
+      "learning_rate": 1.312273716004658e-05,
+      "loss": 0.735,
+      "step": 4099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5893081455973872,
+      "learning_rate": 1.3119607283816823e-05,
+      "loss": 0.7949,
+      "step": 4100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5451259765624377,
+      "learning_rate": 1.3116477069002483e-05,
+      "loss": 0.7979,
+      "step": 4101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6486244621849588,
+      "learning_rate": 1.3113346515943292e-05,
+      "loss": 0.7615,
+      "step": 4102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8091766023002607,
+      "learning_rate": 1.3110215624979026e-05,
+      "loss": 0.7761,
+      "step": 4103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6659017005621068,
+      "learning_rate": 1.3107084396449488e-05,
+      "loss": 0.768,
+      "step": 4104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6204191281973657,
+      "learning_rate": 1.3103952830694531e-05,
+      "loss": 0.8416,
+      "step": 4105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5668821827815804,
+      "learning_rate": 1.3100820928054034e-05,
+      "loss": 0.7551,
+      "step": 4106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.643785128436163,
+      "learning_rate": 1.3097688688867917e-05,
+      "loss": 0.899,
+      "step": 4107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4816697433407158,
+      "learning_rate": 1.3094556113476138e-05,
+      "loss": 0.7447,
+      "step": 4108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3628423656563957,
+      "learning_rate": 1.309142320221868e-05,
+      "loss": 0.7281,
+      "step": 4109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.493882064536198,
+      "learning_rate": 1.3088289955435583e-05,
+      "loss": 0.7529,
+      "step": 4110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5185950409449684,
+      "learning_rate": 1.3085156373466905e-05,
+      "loss": 0.6877,
+      "step": 4111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.552483946978183,
+      "learning_rate": 1.3082022456652748e-05,
+      "loss": 0.619,
+      "step": 4112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6203458777816349,
+      "learning_rate": 1.3078888205333251e-05,
+      "loss": 0.7527,
+      "step": 4113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6555482537000537,
+      "learning_rate": 1.307575361984859e-05,
+      "loss": 0.8066,
+      "step": 4114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.561415647287691,
+      "learning_rate": 1.307261870053897e-05,
+      "loss": 0.7591,
+      "step": 4115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5575464787991349,
+      "learning_rate": 1.3069483447744644e-05,
+      "loss": 0.7045,
+      "step": 4116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.601909026948332,
+      "learning_rate": 1.306634786180589e-05,
+      "loss": 0.7526,
+      "step": 4117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4804890751107598,
+      "learning_rate": 1.3063211943063028e-05,
+      "loss": 0.7148,
+      "step": 4118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4231408741569105,
+      "learning_rate": 1.3060075691856408e-05,
+      "loss": 0.7549,
+      "step": 4119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6377303109941803,
+      "learning_rate": 1.3056939108526432e-05,
+      "loss": 0.8093,
+      "step": 4120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4810282169201698,
+      "learning_rate": 1.3053802193413518e-05,
+      "loss": 0.6746,
+      "step": 4121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.586189039013135,
+      "learning_rate": 1.3050664946858132e-05,
+      "loss": 0.7905,
+      "step": 4122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5164378866528998,
+      "learning_rate": 1.3047527369200774e-05,
+      "loss": 0.703,
+      "step": 4123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.652431828924487,
+      "learning_rate": 1.304438946078198e-05,
+      "loss": 0.7871,
+      "step": 4124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5698746425610295,
+      "learning_rate": 1.3041251221942316e-05,
+      "loss": 0.7083,
+      "step": 4125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.550651917500886,
+      "learning_rate": 1.3038112653022394e-05,
+      "loss": 0.8533,
+      "step": 4126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6512048825368992,
+      "learning_rate": 1.3034973754362851e-05,
+      "loss": 0.9085,
+      "step": 4127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4969760442828155,
+      "learning_rate": 1.303183452630437e-05,
+      "loss": 0.7446,
+      "step": 4128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5726190460076614,
+      "learning_rate": 1.3028694969187665e-05,
+      "loss": 0.743,
+      "step": 4129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4195619694904704,
+      "learning_rate": 1.3025555083353485e-05,
+      "loss": 0.7286,
+      "step": 4130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4969874031232395,
+      "learning_rate": 1.3022414869142612e-05,
+      "loss": 0.6673,
+      "step": 4131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5093059970994227,
+      "learning_rate": 1.3019274326895874e-05,
+      "loss": 0.7385,
+      "step": 4132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5271500356756018,
+      "learning_rate": 1.3016133456954123e-05,
+      "loss": 0.7508,
+      "step": 4133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.520571634239225,
+      "learning_rate": 1.301299225965825e-05,
+      "loss": 0.7333,
+      "step": 4134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5627731896035457,
+      "learning_rate": 1.300985073534919e-05,
+      "loss": 0.7891,
+      "step": 4135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5326098612589172,
+      "learning_rate": 1.3006708884367895e-05,
+      "loss": 0.7618,
+      "step": 4136
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8468622614861878,
+      "learning_rate": 1.3003566707055375e-05,
+      "loss": 0.7747,
+      "step": 4137
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4174643847168273,
+      "learning_rate": 1.3000424203752658e-05,
+      "loss": 0.694,
+      "step": 4138
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4514857062771434,
+      "learning_rate": 1.2997281374800817e-05,
+      "loss": 0.6628,
+      "step": 4139
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5595809922253188,
+      "learning_rate": 1.299413822054095e-05,
+      "loss": 0.85,
+      "step": 4140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4476139519602225,
+      "learning_rate": 1.2990994741314204e-05,
+      "loss": 0.7068,
+      "step": 4141
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5251837298148312,
+      "learning_rate": 1.2987850937461757e-05,
+      "loss": 0.7197,
+      "step": 4142
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4801726487169606,
+      "learning_rate": 1.2984706809324812e-05,
+      "loss": 0.7069,
+      "step": 4143
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5355823466076746,
+      "learning_rate": 1.298156235724462e-05,
+      "loss": 0.684,
+      "step": 4144
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7080840468328307,
+      "learning_rate": 1.2978417581562456e-05,
+      "loss": 0.8024,
+      "step": 4145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5326888897079554,
+      "learning_rate": 1.2975272482619645e-05,
+      "loss": 0.7033,
+      "step": 4146
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.52255665500569,
+      "learning_rate": 1.2972127060757536e-05,
+      "loss": 0.7515,
+      "step": 4147
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6721411657315013,
+      "learning_rate": 1.296898131631751e-05,
+      "loss": 0.8513,
+      "step": 4148
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4392424729686444,
+      "learning_rate": 1.296583524964099e-05,
+      "loss": 0.7486,
+      "step": 4149
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.579784131290916,
+      "learning_rate": 1.296268886106944e-05,
+      "loss": 0.7839,
+      "step": 4150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.442856162503546,
+      "learning_rate": 1.295954215094434e-05,
+      "loss": 0.7541,
+      "step": 4151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5161520779023678,
+      "learning_rate": 1.2956395119607225e-05,
+      "loss": 0.6876,
+      "step": 4152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6270966963220517,
+      "learning_rate": 1.2953247767399653e-05,
+      "loss": 0.7555,
+      "step": 4153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4074602171467065,
+      "learning_rate": 1.2950100094663216e-05,
+      "loss": 0.6801,
+      "step": 4154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.485430352222668,
+      "learning_rate": 1.294695210173955e-05,
+      "loss": 0.6834,
+      "step": 4155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5971365760873109,
+      "learning_rate": 1.2943803788970319e-05,
+      "loss": 0.6976,
+      "step": 4156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5877675607054489,
+      "learning_rate": 1.294065515669722e-05,
+      "loss": 0.7516,
+      "step": 4157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5611545308976658,
+      "learning_rate": 1.2937506205261991e-05,
+      "loss": 0.7916,
+      "step": 4158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5146360432539503,
+      "learning_rate": 1.2934356935006405e-05,
+      "loss": 0.7435,
+      "step": 4159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5328419599085918,
+      "learning_rate": 1.293120734627226e-05,
+      "loss": 0.6888,
+      "step": 4160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6270060946823581,
+      "learning_rate": 1.2928057439401396e-05,
+      "loss": 0.6567,
+      "step": 4161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5686762641960714,
+      "learning_rate": 1.2924907214735686e-05,
+      "loss": 0.7013,
+      "step": 4162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.679353724413454,
+      "learning_rate": 1.2921756672617039e-05,
+      "loss": 0.8016,
+      "step": 4163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7116083356462308,
+      "learning_rate": 1.2918605813387395e-05,
+      "loss": 0.8348,
+      "step": 4164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3332388029642228,
+      "learning_rate": 1.2915454637388733e-05,
+      "loss": 0.6438,
+      "step": 4165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5179907748585992,
+      "learning_rate": 1.2912303144963062e-05,
+      "loss": 0.7783,
+      "step": 4166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5280775014974233,
+      "learning_rate": 1.290915133645243e-05,
+      "loss": 0.7175,
+      "step": 4167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6248733792262533,
+      "learning_rate": 1.2905999212198911e-05,
+      "loss": 0.8058,
+      "step": 4168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5849368736277467,
+      "learning_rate": 1.2902846772544625e-05,
+      "loss": 0.7131,
+      "step": 4169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4466438034909208,
+      "learning_rate": 1.2899694017831717e-05,
+      "loss": 0.6041,
+      "step": 4170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4385438213485517,
+      "learning_rate": 1.289654094840237e-05,
+      "loss": 0.6044,
+      "step": 4171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5468408761996495,
+      "learning_rate": 1.2893387564598798e-05,
+      "loss": 0.7942,
+      "step": 4172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6063767743257966,
+      "learning_rate": 1.2890233866763253e-05,
+      "loss": 0.7492,
+      "step": 4173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5397754358691504,
+      "learning_rate": 1.2887079855238023e-05,
+      "loss": 0.7214,
+      "step": 4174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5545908044398078,
+      "learning_rate": 1.2883925530365422e-05,
+      "loss": 0.7593,
+      "step": 4175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.669074462325384,
+      "learning_rate": 1.2880770892487801e-05,
+      "loss": 0.7603,
+      "step": 4176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.564607040217321,
+      "learning_rate": 1.2877615941947554e-05,
+      "loss": 0.723,
+      "step": 4177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4498352227742795,
+      "learning_rate": 1.2874460679087097e-05,
+      "loss": 0.7274,
+      "step": 4178
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8793408935895344,
+      "learning_rate": 1.2871305104248885e-05,
+      "loss": 0.7631,
+      "step": 4179
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5812346101843897,
+      "learning_rate": 1.2868149217775405e-05,
+      "loss": 0.8137,
+      "step": 4180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.521026332929309,
+      "learning_rate": 1.286499302000918e-05,
+      "loss": 0.705,
+      "step": 4181
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5306191672334164,
+      "learning_rate": 1.2861836511292766e-05,
+      "loss": 0.6851,
+      "step": 4182
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5073339226386875,
+      "learning_rate": 1.2858679691968754e-05,
+      "loss": 0.7127,
+      "step": 4183
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.564368783420573,
+      "learning_rate": 1.2855522562379768e-05,
+      "loss": 0.7618,
+      "step": 4184
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4856689898896016,
+      "learning_rate": 1.2852365122868458e-05,
+      "loss": 0.5892,
+      "step": 4185
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5290378168110395,
+      "learning_rate": 1.2849207373777523e-05,
+      "loss": 0.8697,
+      "step": 4186
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6287002115251614,
+      "learning_rate": 1.2846049315449687e-05,
+      "loss": 0.7245,
+      "step": 4187
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5121503034940078,
+      "learning_rate": 1.2842890948227702e-05,
+      "loss": 0.7317,
+      "step": 4188
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.716114438295996,
+      "learning_rate": 1.2839732272454364e-05,
+      "loss": 0.7015,
+      "step": 4189
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4890403805664274,
+      "learning_rate": 1.283657328847249e-05,
+      "loss": 0.7445,
+      "step": 4190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.604534877043379,
+      "learning_rate": 1.2833413996624953e-05,
+      "loss": 0.73,
+      "step": 4191
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.661337652846229,
+      "learning_rate": 1.2830254397254631e-05,
+      "loss": 0.7738,
+      "step": 4192
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.522246190980118,
+      "learning_rate": 1.2827094490704457e-05,
+      "loss": 0.6997,
+      "step": 4193
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.607181524710579,
+      "learning_rate": 1.2823934277317385e-05,
+      "loss": 0.8442,
+      "step": 4194
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5441419295022962,
+      "learning_rate": 1.2820773757436408e-05,
+      "loss": 0.7235,
+      "step": 4195
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4943906211090183,
+      "learning_rate": 1.2817612931404554e-05,
+      "loss": 0.7262,
+      "step": 4196
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5578734075719667,
+      "learning_rate": 1.281445179956488e-05,
+      "loss": 0.7286,
+      "step": 4197
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.444465296199738,
+      "learning_rate": 1.2811290362260475e-05,
+      "loss": 0.7434,
+      "step": 4198
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3774053157762296,
+      "learning_rate": 1.280812861983446e-05,
+      "loss": 0.6669,
+      "step": 4199
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6867537085739222,
+      "learning_rate": 1.2804966572630004e-05,
+      "loss": 0.718,
+      "step": 4200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7171504576021996,
+      "learning_rate": 1.280180422099029e-05,
+      "loss": 0.7925,
+      "step": 4201
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5436301822788967,
+      "learning_rate": 1.2798641565258542e-05,
+      "loss": 0.7213,
+      "step": 4202
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5605055613090992,
+      "learning_rate": 1.2795478605778016e-05,
+      "loss": 0.6631,
+      "step": 4203
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5283818360565202,
+      "learning_rate": 1.2792315342892007e-05,
+      "loss": 0.7638,
+      "step": 4204
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7239811766109319,
+      "learning_rate": 1.2789151776943833e-05,
+      "loss": 0.8045,
+      "step": 4205
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6300973652859465,
+      "learning_rate": 1.2785987908276853e-05,
+      "loss": 0.6882,
+      "step": 4206
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4571204296952587,
+      "learning_rate": 1.2782823737234452e-05,
+      "loss": 0.7022,
+      "step": 4207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5581818868739223,
+      "learning_rate": 1.2779659264160052e-05,
+      "loss": 0.7965,
+      "step": 4208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5640531713720136,
+      "learning_rate": 1.2776494489397106e-05,
+      "loss": 0.7584,
+      "step": 4209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5477868655715468,
+      "learning_rate": 1.2773329413289108e-05,
+      "loss": 0.6989,
+      "step": 4210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5960710113205359,
+      "learning_rate": 1.2770164036179569e-05,
+      "loss": 0.7414,
+      "step": 4211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4883667796547893,
+      "learning_rate": 1.2766998358412044e-05,
+      "loss": 0.7878,
+      "step": 4212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5053209268600225,
+      "learning_rate": 1.2763832380330118e-05,
+      "loss": 0.7385,
+      "step": 4213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.365287466500688,
+      "learning_rate": 1.276066610227741e-05,
+      "loss": 0.7644,
+      "step": 4214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.617074598757576,
+      "learning_rate": 1.275749952459757e-05,
+      "loss": 0.8595,
+      "step": 4215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5454203983494088,
+      "learning_rate": 1.2754332647634278e-05,
+      "loss": 0.7681,
+      "step": 4216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5272545256382615,
+      "learning_rate": 1.2751165471731248e-05,
+      "loss": 0.6502,
+      "step": 4217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4746003923918978,
+      "learning_rate": 1.274799799723223e-05,
+      "loss": 0.7119,
+      "step": 4218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4663198618339883,
+      "learning_rate": 1.2744830224481004e-05,
+      "loss": 0.6739,
+      "step": 4219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4428619255434645,
+      "learning_rate": 1.2741662153821381e-05,
+      "loss": 0.7681,
+      "step": 4220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4664268102529439,
+      "learning_rate": 1.2738493785597208e-05,
+      "loss": 0.6864,
+      "step": 4221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5853029515488561,
+      "learning_rate": 1.2735325120152357e-05,
+      "loss": 0.7223,
+      "step": 4222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5644251327546417,
+      "learning_rate": 1.2732156157830744e-05,
+      "loss": 0.6766,
+      "step": 4223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4276214621906658,
+      "learning_rate": 1.27289868989763e-05,
+      "loss": 0.6797,
+      "step": 4224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4727346588567345,
+      "learning_rate": 1.272581734393301e-05,
+      "loss": 0.7309,
+      "step": 4225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5771596386172593,
+      "learning_rate": 1.2722647493044873e-05,
+      "loss": 0.779,
+      "step": 4226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.540483221373571,
+      "learning_rate": 1.2719477346655926e-05,
+      "loss": 0.7378,
+      "step": 4227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6694464546919359,
+      "learning_rate": 1.2716306905110243e-05,
+      "loss": 0.7269,
+      "step": 4228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6459659204994768,
+      "learning_rate": 1.2713136168751923e-05,
+      "loss": 0.682,
+      "step": 4229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4297368947921283,
+      "learning_rate": 1.27099651379251e-05,
+      "loss": 0.6766,
+      "step": 4230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5360672121688423,
+      "learning_rate": 1.2706793812973941e-05,
+      "loss": 0.7731,
+      "step": 4231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6790490187862654,
+      "learning_rate": 1.2703622194242644e-05,
+      "loss": 0.7898,
+      "step": 4232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6940426223260903,
+      "learning_rate": 1.2700450282075439e-05,
+      "loss": 0.7994,
+      "step": 4233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6035809726333352,
+      "learning_rate": 1.2697278076816584e-05,
+      "loss": 0.8182,
+      "step": 4234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.911690966187699,
+      "learning_rate": 1.2694105578810372e-05,
+      "loss": 0.7525,
+      "step": 4235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5921485670353903,
+      "learning_rate": 1.2690932788401135e-05,
+      "loss": 0.7356,
+      "step": 4236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.68167081984641,
+      "learning_rate": 1.2687759705933225e-05,
+      "loss": 0.7508,
+      "step": 4237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6075358036783265,
+      "learning_rate": 1.2684586331751027e-05,
+      "loss": 0.734,
+      "step": 4238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4133034871638013,
+      "learning_rate": 1.2681412666198967e-05,
+      "loss": 0.7107,
+      "step": 4239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4271432982156649,
+      "learning_rate": 1.2678238709621495e-05,
+      "loss": 0.6239,
+      "step": 4240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4874506931674698,
+      "learning_rate": 1.2675064462363095e-05,
+      "loss": 0.7471,
+      "step": 4241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6615279803875191,
+      "learning_rate": 1.267188992476828e-05,
+      "loss": 0.7187,
+      "step": 4242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7055012784103516,
+      "learning_rate": 1.2668715097181598e-05,
+      "loss": 0.7781,
+      "step": 4243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5392643779564632,
+      "learning_rate": 1.2665539979947625e-05,
+      "loss": 0.8275,
+      "step": 4244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6392295533469456,
+      "learning_rate": 1.2662364573410972e-05,
+      "loss": 0.7079,
+      "step": 4245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.467439358859539,
+      "learning_rate": 1.265918887791628e-05,
+      "loss": 0.8107,
+      "step": 4246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5340914320267773,
+      "learning_rate": 1.265601289380822e-05,
+      "loss": 0.6401,
+      "step": 4247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6326046668917402,
+      "learning_rate": 1.2652836621431494e-05,
+      "loss": 0.6579,
+      "step": 4248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6060006722081543,
+      "learning_rate": 1.264966006113084e-05,
+      "loss": 0.7607,
+      "step": 4249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6301556197140081,
+      "learning_rate": 1.2646483213251024e-05,
+      "loss": 0.7725,
+      "step": 4250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4477180043687317,
+      "learning_rate": 1.2643306078136839e-05,
+      "loss": 0.6654,
+      "step": 4251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7183853097079613,
+      "learning_rate": 1.264012865613312e-05,
+      "loss": 0.757,
+      "step": 4252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4991277481962177,
+      "learning_rate": 1.2636950947584716e-05,
+      "loss": 0.7827,
+      "step": 4253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3682709122047572,
+      "learning_rate": 1.2633772952836527e-05,
+      "loss": 0.6954,
+      "step": 4254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5264634061223141,
+      "learning_rate": 1.2630594672233474e-05,
+      "loss": 0.6219,
+      "step": 4255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6081603775260733,
+      "learning_rate": 1.2627416106120505e-05,
+      "loss": 0.7464,
+      "step": 4256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6664012905975172,
+      "learning_rate": 1.2624237254842606e-05,
+      "loss": 0.8268,
+      "step": 4257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7236809343198751,
+      "learning_rate": 1.2621058118744789e-05,
+      "loss": 0.7237,
+      "step": 4258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7199094659192433,
+      "learning_rate": 1.2617878698172106e-05,
+      "loss": 0.8653,
+      "step": 4259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5702803700581136,
+      "learning_rate": 1.2614698993469627e-05,
+      "loss": 0.7597,
+      "step": 4260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6599947985801582,
+      "learning_rate": 1.2611519004982463e-05,
+      "loss": 0.72,
+      "step": 4261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5007726034218898,
+      "learning_rate": 1.2608338733055746e-05,
+      "loss": 0.7291,
+      "step": 4262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6180141696571795,
+      "learning_rate": 1.2605158178034656e-05,
+      "loss": 0.8082,
+      "step": 4263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5370907818119977,
+      "learning_rate": 1.2601977340264385e-05,
+      "loss": 0.7279,
+      "step": 4264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6017277705458968,
+      "learning_rate": 1.2598796220090162e-05,
+      "loss": 0.7468,
+      "step": 4265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.470357468363117,
+      "learning_rate": 1.259561481785725e-05,
+      "loss": 0.7744,
+      "step": 4266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5011085967488755,
+      "learning_rate": 1.2592433133910937e-05,
+      "loss": 0.7894,
+      "step": 4267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8058621580428456,
+      "learning_rate": 1.2589251168596554e-05,
+      "loss": 0.6833,
+      "step": 4268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5790531507207646,
+      "learning_rate": 1.2586068922259446e-05,
+      "loss": 0.807,
+      "step": 4269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6129616607668176,
+      "learning_rate": 1.2582886395245e-05,
+      "loss": 0.7582,
+      "step": 4270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5636248037220053,
+      "learning_rate": 1.2579703587898622e-05,
+      "loss": 0.7492,
+      "step": 4271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6477268242323033,
+      "learning_rate": 1.2576520500565767e-05,
+      "loss": 0.8129,
+      "step": 4272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.53427061103123,
+      "learning_rate": 1.2573337133591904e-05,
+      "loss": 0.7485,
+      "step": 4273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2760812692182046,
+      "learning_rate": 1.2570153487322534e-05,
+      "loss": 0.6305,
+      "step": 4274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4831290425463768,
+      "learning_rate": 1.2566969562103197e-05,
+      "loss": 0.6692,
+      "step": 4275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4653955773573912,
+      "learning_rate": 1.2563785358279459e-05,
+      "loss": 0.6619,
+      "step": 4276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3376784691184405,
+      "learning_rate": 1.2560600876196911e-05,
+      "loss": 0.6905,
+      "step": 4277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5494933539636409,
+      "learning_rate": 1.255741611620118e-05,
+      "loss": 0.7156,
+      "step": 4278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8980840079305699,
+      "learning_rate": 1.2554231078637926e-05,
+      "loss": 0.7758,
+      "step": 4279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5326597254174636,
+      "learning_rate": 1.2551045763852832e-05,
+      "loss": 0.7281,
+      "step": 4280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6493596040522875,
+      "learning_rate": 1.2547860172191613e-05,
+      "loss": 0.8091,
+      "step": 4281
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6204455055952354,
+      "learning_rate": 1.2544674304000017e-05,
+      "loss": 0.7701,
+      "step": 4282
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5362361176412207,
+      "learning_rate": 1.254148815962382e-05,
+      "loss": 0.6546,
+      "step": 4283
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5859203814985514,
+      "learning_rate": 1.2538301739408827e-05,
+      "loss": 0.7461,
+      "step": 4284
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6904326299259982,
+      "learning_rate": 1.2535115043700873e-05,
+      "loss": 0.7867,
+      "step": 4285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.49829461015147,
+      "learning_rate": 1.2531928072845829e-05,
+      "loss": 0.7796,
+      "step": 4286
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5082088395185627,
+      "learning_rate": 1.2528740827189588e-05,
+      "loss": 0.7096,
+      "step": 4287
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6725549597277543,
+      "learning_rate": 1.2525553307078074e-05,
+      "loss": 0.7854,
+      "step": 4288
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4817730777320866,
+      "learning_rate": 1.2522365512857241e-05,
+      "loss": 0.6682,
+      "step": 4289
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.535915561788933,
+      "learning_rate": 1.251917744487308e-05,
+      "loss": 0.7609,
+      "step": 4290
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5623924070870114,
+      "learning_rate": 1.2515989103471602e-05,
+      "loss": 0.6874,
+      "step": 4291
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.525575165734888,
+      "learning_rate": 1.2512800488998856e-05,
+      "loss": 0.8236,
+      "step": 4292
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7352730208482219,
+      "learning_rate": 1.2509611601800908e-05,
+      "loss": 0.6995,
+      "step": 4293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6429301062021924,
+      "learning_rate": 1.2506422442223867e-05,
+      "loss": 0.8077,
+      "step": 4294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4617991681259823,
+      "learning_rate": 1.2503233010613866e-05,
+      "loss": 0.6972,
+      "step": 4295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5839803085209656,
+      "learning_rate": 1.2500043307317072e-05,
+      "loss": 0.7635,
+      "step": 4296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4985757263506088,
+      "learning_rate": 1.2496853332679668e-05,
+      "loss": 0.68,
+      "step": 4297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.876952140170346,
+      "learning_rate": 1.2493663087047883e-05,
+      "loss": 0.8143,
+      "step": 4298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4912297726382522,
+      "learning_rate": 1.2490472570767966e-05,
+      "loss": 0.7566,
+      "step": 4299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4135576446856426,
+      "learning_rate": 1.24872817841862e-05,
+      "loss": 0.7676,
+      "step": 4300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6894482298761555,
+      "learning_rate": 1.2484090727648892e-05,
+      "loss": 0.7346,
+      "step": 4301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.410949297535782,
+      "learning_rate": 1.2480899401502384e-05,
+      "loss": 0.698,
+      "step": 4302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4645565318403775,
+      "learning_rate": 1.247770780609304e-05,
+      "loss": 0.6811,
+      "step": 4303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6854348277084623,
+      "learning_rate": 1.2474515941767262e-05,
+      "loss": 0.7574,
+      "step": 4304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7077740397861376,
+      "learning_rate": 1.2471323808871475e-05,
+      "loss": 0.7648,
+      "step": 4305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.670997861295573,
+      "learning_rate": 1.2468131407752138e-05,
+      "loss": 0.7125,
+      "step": 4306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6537332834926746,
+      "learning_rate": 1.246493873875573e-05,
+      "loss": 0.9453,
+      "step": 4307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6298292725373442,
+      "learning_rate": 1.2461745802228771e-05,
+      "loss": 0.6995,
+      "step": 4308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6336178608807248,
+      "learning_rate": 1.2458552598517804e-05,
+      "loss": 0.9437,
+      "step": 4309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6605536767879754,
+      "learning_rate": 1.24553591279694e-05,
+      "loss": 0.7589,
+      "step": 4310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6214145562153597,
+      "learning_rate": 1.2452165390930159e-05,
+      "loss": 0.756,
+      "step": 4311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7693944283318281,
+      "learning_rate": 1.244897138774671e-05,
+      "loss": 0.776,
+      "step": 4312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9429667641152428,
+      "learning_rate": 1.244577711876572e-05,
+      "loss": 0.7921,
+      "step": 4313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4719654131064106,
+      "learning_rate": 1.2442582584333867e-05,
+      "loss": 0.7164,
+      "step": 4314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6998681620739857,
+      "learning_rate": 1.2439387784797873e-05,
+      "loss": 0.7358,
+      "step": 4315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5367814354024754,
+      "learning_rate": 1.2436192720504483e-05,
+      "loss": 0.7949,
+      "step": 4316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5814752360852835,
+      "learning_rate": 1.2432997391800471e-05,
+      "loss": 0.7588,
+      "step": 4317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5482592457995743,
+      "learning_rate": 1.242980179903264e-05,
+      "loss": 0.7803,
+      "step": 4318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.543902286716255,
+      "learning_rate": 1.2426605942547822e-05,
+      "loss": 0.7623,
+      "step": 4319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5364496774447027,
+      "learning_rate": 1.2423409822692878e-05,
+      "loss": 0.7359,
+      "step": 4320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7470060336826811,
+      "learning_rate": 1.2420213439814693e-05,
+      "loss": 0.8224,
+      "step": 4321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7947145930976374,
+      "learning_rate": 1.2417016794260188e-05,
+      "loss": 0.9604,
+      "step": 4322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4318612043412737,
+      "learning_rate": 1.2413819886376308e-05,
+      "loss": 0.7188,
+      "step": 4323
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.578719074009928,
+      "learning_rate": 1.241062271651003e-05,
+      "loss": 0.7632,
+      "step": 4324
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4222966215231232,
+      "learning_rate": 1.2407425285008348e-05,
+      "loss": 0.727,
+      "step": 4325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7037074418102436,
+      "learning_rate": 1.2404227592218302e-05,
+      "loss": 0.7614,
+      "step": 4326
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5560684096552946,
+      "learning_rate": 1.2401029638486952e-05,
+      "loss": 0.7566,
+      "step": 4327
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5496149562355548,
+      "learning_rate": 1.2397831424161379e-05,
+      "loss": 0.6464,
+      "step": 4328
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5661260611233834,
+      "learning_rate": 1.2394632949588702e-05,
+      "loss": 0.7027,
+      "step": 4329
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.647435871699377,
+      "learning_rate": 1.239143421511607e-05,
+      "loss": 0.7214,
+      "step": 4330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6376179783423268,
+      "learning_rate": 1.238823522109065e-05,
+      "loss": 0.7569,
+      "step": 4331
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7213715524105555,
+      "learning_rate": 1.2385035967859644e-05,
+      "loss": 0.8125,
+      "step": 4332
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5625709746455267,
+      "learning_rate": 1.2381836455770281e-05,
+      "loss": 0.696,
+      "step": 4333
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5850692435817693,
+      "learning_rate": 1.237863668516982e-05,
+      "loss": 0.8419,
+      "step": 4334
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6029512722585655,
+      "learning_rate": 1.2375436656405543e-05,
+      "loss": 0.8223,
+      "step": 4335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4897526621851649,
+      "learning_rate": 1.2372236369824766e-05,
+      "loss": 0.7425,
+      "step": 4336
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4048163565710574,
+      "learning_rate": 1.236903582577483e-05,
+      "loss": 0.6398,
+      "step": 4337
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5031420071231552,
+      "learning_rate": 1.2365835024603099e-05,
+      "loss": 0.8474,
+      "step": 4338
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3533265566661303,
+      "learning_rate": 1.2362633966656974e-05,
+      "loss": 0.5881,
+      "step": 4339
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6080843502117637,
+      "learning_rate": 1.2359432652283882e-05,
+      "loss": 0.7397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6916691645649253,
+      "learning_rate": 1.2356231081831274e-05,
+      "loss": 0.835,
+      "step": 4341
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.541941556966986,
+      "learning_rate": 1.2353029255646627e-05,
+      "loss": 0.7793,
+      "step": 4342
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3431183490118175,
+      "learning_rate": 1.234982717407745e-05,
+      "loss": 0.6793,
+      "step": 4343
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5996485332722739,
+      "learning_rate": 1.2346624837471282e-05,
+      "loss": 0.8009,
+      "step": 4344
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5334897538387262,
+      "learning_rate": 1.2343422246175687e-05,
+      "loss": 0.6806,
+      "step": 4345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5390653761364153,
+      "learning_rate": 1.2340219400538251e-05,
+      "loss": 0.7227,
+      "step": 4346
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5577072860816348,
+      "learning_rate": 1.2337016300906599e-05,
+      "loss": 0.7152,
+      "step": 4347
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4338529859628366,
+      "learning_rate": 1.2333812947628372e-05,
+      "loss": 0.6529,
+      "step": 4348
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5103611554283487,
+      "learning_rate": 1.2330609341051248e-05,
+      "loss": 0.5906,
+      "step": 4349
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5452087394046694,
+      "learning_rate": 1.2327405481522925e-05,
+      "loss": 0.7806,
+      "step": 4350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.506080493851666,
+      "learning_rate": 1.2324201369391135e-05,
+      "loss": 0.7061,
+      "step": 4351
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.611522681979579,
+      "learning_rate": 1.232099700500363e-05,
+      "loss": 0.7171,
+      "step": 4352
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4512234024915196,
+      "learning_rate": 1.23177923887082e-05,
+      "loss": 0.7174,
+      "step": 4353
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4573865532669146,
+      "learning_rate": 1.2314587520852652e-05,
+      "loss": 0.6301,
+      "step": 4354
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.639497091761245,
+      "learning_rate": 1.2311382401784823e-05,
+      "loss": 0.7631,
+      "step": 4355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.502122098326732,
+      "learning_rate": 1.2308177031852581e-05,
+      "loss": 0.7225,
+      "step": 4356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6607915454021849,
+      "learning_rate": 1.2304971411403818e-05,
+      "loss": 0.7495,
+      "step": 4357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4264459129384013,
+      "learning_rate": 1.2301765540786454e-05,
+      "loss": 0.6694,
+      "step": 4358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4441931508298085,
+      "learning_rate": 1.2298559420348437e-05,
+      "loss": 0.6286,
+      "step": 4359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6350542372791959,
+      "learning_rate": 1.229535305043774e-05,
+      "loss": 0.7763,
+      "step": 4360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5076741786633405,
+      "learning_rate": 1.2292146431402364e-05,
+      "loss": 0.7056,
+      "step": 4361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4946756140117825,
+      "learning_rate": 1.2288939563590336e-05,
+      "loss": 0.7498,
+      "step": 4362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4954116089047855,
+      "learning_rate": 1.2285732447349718e-05,
+      "loss": 0.7052,
+      "step": 4363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.539129278892199,
+      "learning_rate": 1.2282525083028585e-05,
+      "loss": 0.7045,
+      "step": 4364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5616495206143877,
+      "learning_rate": 1.227931747097505e-05,
+      "loss": 0.7171,
+      "step": 4365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.541181730746174,
+      "learning_rate": 1.2276109611537246e-05,
+      "loss": 0.6921,
+      "step": 4366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.574039031315957,
+      "learning_rate": 1.2272901505063339e-05,
+      "loss": 0.7405,
+      "step": 4367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6611638774575088,
+      "learning_rate": 1.226969315190152e-05,
+      "loss": 0.7466,
+      "step": 4368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.795872991044649,
+      "learning_rate": 1.2266484552400002e-05,
+      "loss": 0.7829,
+      "step": 4369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5984338918264582,
+      "learning_rate": 1.2263275706907028e-05,
+      "loss": 0.6836,
+      "step": 4370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6082604218617218,
+      "learning_rate": 1.2260066615770872e-05,
+      "loss": 0.7268,
+      "step": 4371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5863700764160853,
+      "learning_rate": 1.2256857279339826e-05,
+      "loss": 0.7271,
+      "step": 4372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.502654497794165,
+      "learning_rate": 1.2253647697962219e-05,
+      "loss": 0.7056,
+      "step": 4373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.455642030024103,
+      "learning_rate": 1.2250437871986396e-05,
+      "loss": 0.6944,
+      "step": 4374
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.732673603633385,
+      "learning_rate": 1.2247227801760733e-05,
+      "loss": 0.8043,
+      "step": 4375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5572231314499265,
+      "learning_rate": 1.2244017487633637e-05,
+      "loss": 0.6938,
+      "step": 4376
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.687209376387509,
+      "learning_rate": 1.2240806929953535e-05,
+      "loss": 0.703,
+      "step": 4377
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5609158173524036,
+      "learning_rate": 1.2237596129068883e-05,
+      "loss": 0.812,
+      "step": 4378
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.530954813512489,
+      "learning_rate": 1.223438508532816e-05,
+      "loss": 0.77,
+      "step": 4379
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5362119262865612,
+      "learning_rate": 1.2231173799079881e-05,
+      "loss": 0.718,
+      "step": 4380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4563236326328328,
+      "learning_rate": 1.2227962270672578e-05,
+      "loss": 0.7434,
+      "step": 4381
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6876187955506705,
+      "learning_rate": 1.2224750500454812e-05,
+      "loss": 0.7205,
+      "step": 4382
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5422731338231965,
+      "learning_rate": 1.2221538488775169e-05,
+      "loss": 0.7501,
+      "step": 4383
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6188008465025745,
+      "learning_rate": 1.2218326235982265e-05,
+      "loss": 0.7096,
+      "step": 4384
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5020069685350754,
+      "learning_rate": 1.2215113742424737e-05,
+      "loss": 0.7346,
+      "step": 4385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5813327112219508,
+      "learning_rate": 1.2211901008451254e-05,
+      "loss": 0.6556,
+      "step": 4386
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.666893881592953,
+      "learning_rate": 1.2208688034410508e-05,
+      "loss": 0.7044,
+      "step": 4387
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5003835672667813,
+      "learning_rate": 1.2205474820651216e-05,
+      "loss": 0.7246,
+      "step": 4388
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.549861347647712,
+      "learning_rate": 1.2202261367522119e-05,
+      "loss": 0.6512,
+      "step": 4389
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5602212910735438,
+      "learning_rate": 1.2199047675371993e-05,
+      "loss": 0.7465,
+      "step": 4390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.753807997087651,
+      "learning_rate": 1.219583374454963e-05,
+      "loss": 0.7398,
+      "step": 4391
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6286670549259223,
+      "learning_rate": 1.2192619575403855e-05,
+      "loss": 0.7511,
+      "step": 4392
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3647542817459297,
+      "learning_rate": 1.218940516828351e-05,
+      "loss": 0.6561,
+      "step": 4393
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5529890621235534,
+      "learning_rate": 1.2186190523537476e-05,
+      "loss": 0.695,
+      "step": 4394
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5510040334936726,
+      "learning_rate": 1.2182975641514646e-05,
+      "loss": 0.6774,
+      "step": 4395
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5926843629490448,
+      "learning_rate": 1.217976052256395e-05,
+      "loss": 0.7498,
+      "step": 4396
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5187507805011788,
+      "learning_rate": 1.2176545167034337e-05,
+      "loss": 0.7326,
+      "step": 4397
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6532601994941343,
+      "learning_rate": 1.2173329575274781e-05,
+      "loss": 0.6889,
+      "step": 4398
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5301841899089985,
+      "learning_rate": 1.217011374763429e-05,
+      "loss": 0.8719,
+      "step": 4399
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5979785297825504,
+      "learning_rate": 1.2166897684461887e-05,
+      "loss": 0.6897,
+      "step": 4400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4936498800599838,
+      "learning_rate": 1.2163681386106628e-05,
+      "loss": 0.7131,
+      "step": 4401
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.561516544759848,
+      "learning_rate": 1.2160464852917586e-05,
+      "loss": 0.7337,
+      "step": 4402
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6101791995119874,
+      "learning_rate": 1.2157248085243875e-05,
+      "loss": 0.7181,
+      "step": 4403
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4056365477488921,
+      "learning_rate": 1.2154031083434618e-05,
+      "loss": 0.7246,
+      "step": 4404
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4791446978711529,
+      "learning_rate": 1.2150813847838973e-05,
+      "loss": 0.7785,
+      "step": 4405
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6116695930351572,
+      "learning_rate": 1.2147596378806122e-05,
+      "loss": 0.8783,
+      "step": 4406
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.525736879948389,
+      "learning_rate": 1.2144378676685265e-05,
+      "loss": 0.6726,
+      "step": 4407
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.704644176726741,
+      "learning_rate": 1.2141160741825637e-05,
+      "loss": 0.8096,
+      "step": 4408
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6201455401565257,
+      "learning_rate": 1.2137942574576498e-05,
+      "loss": 0.7243,
+      "step": 4409
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5974565990510263,
+      "learning_rate": 1.2134724175287124e-05,
+      "loss": 0.7345,
+      "step": 4410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5472920340894925,
+      "learning_rate": 1.213150554430682e-05,
+      "loss": 0.6502,
+      "step": 4411
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6689059927391285,
+      "learning_rate": 1.2128286681984929e-05,
+      "loss": 0.8014,
+      "step": 4412
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7726442768700028,
+      "learning_rate": 1.2125067588670798e-05,
+      "loss": 0.6601,
+      "step": 4413
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6477284044940803,
+      "learning_rate": 1.2121848264713813e-05,
+      "loss": 0.7439,
+      "step": 4414
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4965439315138036,
+      "learning_rate": 1.2118628710463383e-05,
+      "loss": 0.6484,
+      "step": 4415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5165985892615441,
+      "learning_rate": 1.2115408926268935e-05,
+      "loss": 0.7808,
+      "step": 4416
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5558271212569084,
+      "learning_rate": 1.211218891247993e-05,
+      "loss": 0.6624,
+      "step": 4417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5968364577633354,
+      "learning_rate": 1.210896866944585e-05,
+      "loss": 0.7942,
+      "step": 4418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5649442392983401,
+      "learning_rate": 1.2105748197516202e-05,
+      "loss": 0.7289,
+      "step": 4419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4583243392481153,
+      "learning_rate": 1.2102527497040518e-05,
+      "loss": 0.7746,
+      "step": 4420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4272608738704586,
+      "learning_rate": 1.2099306568368351e-05,
+      "loss": 0.645,
+      "step": 4421
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4106183228923768,
+      "learning_rate": 1.2096085411849288e-05,
+      "loss": 0.7285,
+      "step": 4422
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5010831315856243,
+      "learning_rate": 1.2092864027832934e-05,
+      "loss": 0.6688,
+      "step": 4423
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6271030020091606,
+      "learning_rate": 1.2089642416668917e-05,
+      "loss": 0.7768,
+      "step": 4424
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5235151617778604,
+      "learning_rate": 1.2086420578706892e-05,
+      "loss": 0.6403,
+      "step": 4425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5076423948411202,
+      "learning_rate": 1.2083198514296543e-05,
+      "loss": 0.7494,
+      "step": 4426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5464772791066064,
+      "learning_rate": 1.2079976223787572e-05,
+      "loss": 0.889,
+      "step": 4427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5354979787731868,
+      "learning_rate": 1.2076753707529712e-05,
+      "loss": 0.6382,
+      "step": 4428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.423218504781596,
+      "learning_rate": 1.207353096587271e-05,
+      "loss": 0.7312,
+      "step": 4429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3970447157524866,
+      "learning_rate": 1.2070307999166349e-05,
+      "loss": 0.6877,
+      "step": 4430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6599140309374598,
+      "learning_rate": 1.2067084807760432e-05,
+      "loss": 0.7027,
+      "step": 4431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6651396370667126,
+      "learning_rate": 1.2063861392004786e-05,
+      "loss": 0.7474,
+      "step": 4432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.430089832277469,
+      "learning_rate": 1.2060637752249258e-05,
+      "loss": 0.7243,
+      "step": 4433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.614481881170845,
+      "learning_rate": 1.2057413888843725e-05,
+      "loss": 0.8231,
+      "step": 4434
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4348838527996164,
+      "learning_rate": 1.2054189802138092e-05,
+      "loss": 0.7713,
+      "step": 4435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7212586577126978,
+      "learning_rate": 1.205096549248228e-05,
+      "loss": 0.658,
+      "step": 4436
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.658407315132644,
+      "learning_rate": 1.2047740960226237e-05,
+      "loss": 0.7476,
+      "step": 4437
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5328008396841144,
+      "learning_rate": 1.2044516205719936e-05,
+      "loss": 0.8208,
+      "step": 4438
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4121317489710041,
+      "learning_rate": 1.2041291229313372e-05,
+      "loss": 0.7737,
+      "step": 4439
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6338576991003722,
+      "learning_rate": 1.2038066031356568e-05,
+      "loss": 0.777,
+      "step": 4440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6169747349261165,
+      "learning_rate": 1.203484061219957e-05,
+      "loss": 0.7694,
+      "step": 4441
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.534574166088398,
+      "learning_rate": 1.2031614972192443e-05,
+      "loss": 0.7543,
+      "step": 4442
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5235420808364872,
+      "learning_rate": 1.2028389111685283e-05,
+      "loss": 0.766,
+      "step": 4443
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.505387947763993,
+      "learning_rate": 1.2025163031028203e-05,
+      "loss": 0.8059,
+      "step": 4444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5351796430514244,
+      "learning_rate": 1.202193673057135e-05,
+      "loss": 0.6711,
+      "step": 4445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5112091190626977,
+      "learning_rate": 1.201871021066488e-05,
+      "loss": 0.8201,
+      "step": 4446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.627302058235011,
+      "learning_rate": 1.2015483471658986e-05,
+      "loss": 0.7077,
+      "step": 4447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0278561697736004,
+      "learning_rate": 1.2012256513903881e-05,
+      "loss": 0.7857,
+      "step": 4448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.611940844306635,
+      "learning_rate": 1.2009029337749803e-05,
+      "loss": 0.7612,
+      "step": 4449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8121778869497533,
+      "learning_rate": 1.2005801943547004e-05,
+      "loss": 0.78,
+      "step": 4450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5552713757449876,
+      "learning_rate": 1.2002574331645774e-05,
+      "loss": 0.7561,
+      "step": 4451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6294178143747788,
+      "learning_rate": 1.1999346502396411e-05,
+      "loss": 0.6489,
+      "step": 4452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5805010400813575,
+      "learning_rate": 1.1996118456149258e-05,
+      "loss": 0.6256,
+      "step": 4453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6173180486478613,
+      "learning_rate": 1.199289019325466e-05,
+      "loss": 0.7834,
+      "step": 4454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4599722225184395,
+      "learning_rate": 1.1989661714063e-05,
+      "loss": 0.7161,
+      "step": 4455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5629592431617074,
+      "learning_rate": 1.198643301892467e-05,
+      "loss": 0.6841,
+      "step": 4456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6195847291825776,
+      "learning_rate": 1.1983204108190102e-05,
+      "loss": 0.7359,
+      "step": 4457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6672599164503639,
+      "learning_rate": 1.1979974982209745e-05,
+      "loss": 0.8227,
+      "step": 4458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6956380914634728,
+      "learning_rate": 1.1976745641334064e-05,
+      "loss": 0.7724,
+      "step": 4459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3448791245225744,
+      "learning_rate": 1.197351608591356e-05,
+      "loss": 0.6883,
+      "step": 4460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4919968482707506,
+      "learning_rate": 1.1970286316298743e-05,
+      "loss": 0.8661,
+      "step": 4461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5888723787256789,
+      "learning_rate": 1.1967056332840159e-05,
+      "loss": 0.6612,
+      "step": 4462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.500626958840491,
+      "learning_rate": 1.1963826135888374e-05,
+      "loss": 0.6686,
+      "step": 4463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3928861170099849,
+      "learning_rate": 1.196059572579397e-05,
+      "loss": 0.7959,
+      "step": 4464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6296766789886576,
+      "learning_rate": 1.1957365102907562e-05,
+      "loss": 0.8137,
+      "step": 4465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0328265489817867,
+      "learning_rate": 1.1954134267579781e-05,
+      "loss": 0.8245,
+      "step": 4466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.654234354125284,
+      "learning_rate": 1.1950903220161286e-05,
+      "loss": 0.6642,
+      "step": 4467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5800488068146095,
+      "learning_rate": 1.1947671961002753e-05,
+      "loss": 0.7177,
+      "step": 4468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4788972818110588,
+      "learning_rate": 1.1944440490454887e-05,
+      "loss": 0.764,
+      "step": 4469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7483788792193522,
+      "learning_rate": 1.1941208808868411e-05,
+      "loss": 0.7508,
+      "step": 4470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.518687096393544,
+      "learning_rate": 1.193797691659408e-05,
+      "loss": 0.8022,
+      "step": 4471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5851103825987014,
+      "learning_rate": 1.1934744813982662e-05,
+      "loss": 0.7285,
+      "step": 4472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5147747934349225,
+      "learning_rate": 1.193151250138495e-05,
+      "loss": 0.6713,
+      "step": 4473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7252009780168112,
+      "learning_rate": 1.1928279979151757e-05,
+      "loss": 0.8059,
+      "step": 4474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6550933483637722,
+      "learning_rate": 1.1925047247633931e-05,
+      "loss": 0.7824,
+      "step": 4475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.537864083464159,
+      "learning_rate": 1.1921814307182331e-05,
+      "loss": 0.7198,
+      "step": 4476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6247694014405787,
+      "learning_rate": 1.1918581158147842e-05,
+      "loss": 0.7996,
+      "step": 4477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5040273170796579,
+      "learning_rate": 1.1915347800881369e-05,
+      "loss": 0.7318,
+      "step": 4478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.624008872920982,
+      "learning_rate": 1.1912114235733846e-05,
+      "loss": 0.8635,
+      "step": 4479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6783046655630303,
+      "learning_rate": 1.1908880463056225e-05,
+      "loss": 0.7072,
+      "step": 4480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5987382924783546,
+      "learning_rate": 1.1905646483199481e-05,
+      "loss": 0.737,
+      "step": 4481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4190839686862713,
+      "learning_rate": 1.1902412296514614e-05,
+      "loss": 0.6069,
+      "step": 4482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4717341097697567,
+      "learning_rate": 1.1899177903352639e-05,
+      "loss": 0.6212,
+      "step": 4483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5764180668532193,
+      "learning_rate": 1.1895943304064606e-05,
+      "loss": 0.6834,
+      "step": 4484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.545690402929286,
+      "learning_rate": 1.1892708499001576e-05,
+      "loss": 0.613,
+      "step": 4485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5153286028527162,
+      "learning_rate": 1.1889473488514636e-05,
+      "loss": 0.7848,
+      "step": 4486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5809882519403804,
+      "learning_rate": 1.1886238272954897e-05,
+      "loss": 0.7388,
+      "step": 4487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5637085518869727,
+      "learning_rate": 1.1883002852673489e-05,
+      "loss": 0.7386,
+      "step": 4488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.522366166224289,
+      "learning_rate": 1.187976722802157e-05,
+      "loss": 0.6202,
+      "step": 4489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6818519644180367,
+      "learning_rate": 1.1876531399350316e-05,
+      "loss": 0.7759,
+      "step": 4490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.552904746172006,
+      "learning_rate": 1.1873295367010923e-05,
+      "loss": 0.7709,
+      "step": 4491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4481819506527602,
+      "learning_rate": 1.1870059131354611e-05,
+      "loss": 0.7444,
+      "step": 4492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5800787460222114,
+      "learning_rate": 1.186682269273263e-05,
+      "loss": 0.8605,
+      "step": 4493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6270808444334217,
+      "learning_rate": 1.1863586051496235e-05,
+      "loss": 0.6664,
+      "step": 4494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3324751236800856,
+      "learning_rate": 1.1860349207996717e-05,
+      "loss": 0.7568,
+      "step": 4495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6101269155158606,
+      "learning_rate": 1.1857112162585387e-05,
+      "loss": 0.811,
+      "step": 4496
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5018243352019984,
+      "learning_rate": 1.1853874915613575e-05,
+      "loss": 0.816,
+      "step": 4497
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6115192399287763,
+      "learning_rate": 1.1850637467432629e-05,
+      "loss": 0.7215,
+      "step": 4498
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5961199154311783,
+      "learning_rate": 1.184739981839393e-05,
+      "loss": 0.6597,
+      "step": 4499
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.577479387062586,
+      "learning_rate": 1.1844161968848868e-05,
+      "loss": 0.7565,
+      "step": 4500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5304523479785352,
+      "learning_rate": 1.1840923919148863e-05,
+      "loss": 0.6509,
+      "step": 4501
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.318352712290105,
+      "learning_rate": 1.1837685669645358e-05,
+      "loss": 0.6931,
+      "step": 4502
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5895914554412998,
+      "learning_rate": 1.1834447220689813e-05,
+      "loss": 0.737,
+      "step": 4503
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5971370797130637,
+      "learning_rate": 1.1831208572633708e-05,
+      "loss": 0.7396,
+      "step": 4504
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.544909155827977,
+      "learning_rate": 1.182796972582855e-05,
+      "loss": 0.6391,
+      "step": 4505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5076713528789134,
+      "learning_rate": 1.1824730680625862e-05,
+      "loss": 0.7217,
+      "step": 4506
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6015860072889743,
+      "learning_rate": 1.1821491437377198e-05,
+      "loss": 0.7712,
+      "step": 4507
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.559399696293891,
+      "learning_rate": 1.181825199643412e-05,
+      "loss": 0.682,
+      "step": 4508
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3669347461286494,
+      "learning_rate": 1.1815012358148223e-05,
+      "loss": 0.5674,
+      "step": 4509
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7240826706457286,
+      "learning_rate": 1.1811772522871119e-05,
+      "loss": 0.7303,
+      "step": 4510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6573139580065175,
+      "learning_rate": 1.1808532490954438e-05,
+      "loss": 0.7783,
+      "step": 4511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6018607990043154,
+      "learning_rate": 1.180529226274984e-05,
+      "loss": 0.8296,
+      "step": 4512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6779980575994735,
+      "learning_rate": 1.1802051838609e-05,
+      "loss": 0.7266,
+      "step": 4513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4748993331800264,
+      "learning_rate": 1.1798811218883613e-05,
+      "loss": 0.7311,
+      "step": 4514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6332739658424689,
+      "learning_rate": 1.1795570403925397e-05,
+      "loss": 0.6996,
+      "step": 4515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.520362224377642,
+      "learning_rate": 1.1792329394086094e-05,
+      "loss": 0.6282,
+      "step": 4516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3075697117316751,
+      "learning_rate": 1.1789088189717463e-05,
+      "loss": 0.7129,
+      "step": 4517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6161533444931127,
+      "learning_rate": 1.178584679117129e-05,
+      "loss": 0.6684,
+      "step": 4518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5433187962870167,
+      "learning_rate": 1.1782605198799371e-05,
+      "loss": 0.721,
+      "step": 4519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4472065593881311,
+      "learning_rate": 1.1779363412953537e-05,
+      "loss": 0.806,
+      "step": 4520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4993530836356177,
+      "learning_rate": 1.177612143398563e-05,
+      "loss": 0.7325,
+      "step": 4521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7396800214504256,
+      "learning_rate": 1.1772879262247519e-05,
+      "loss": 0.7748,
+      "step": 4522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7334103979363331,
+      "learning_rate": 1.1769636898091089e-05,
+      "loss": 0.6951,
+      "step": 4523
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5340003104618167,
+      "learning_rate": 1.1766394341868242e-05,
+      "loss": 0.6365,
+      "step": 4524
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5822547649750762,
+      "learning_rate": 1.1763151593930918e-05,
+      "loss": 0.7985,
+      "step": 4525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4388485063230003,
+      "learning_rate": 1.1759908654631057e-05,
+      "loss": 0.7047,
+      "step": 4526
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6219002551817887,
+      "learning_rate": 1.1756665524320638e-05,
+      "loss": 0.6781,
+      "step": 4527
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5640124456440354,
+      "learning_rate": 1.1753422203351643e-05,
+      "loss": 0.7798,
+      "step": 4528
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3788120020476304,
+      "learning_rate": 1.1750178692076092e-05,
+      "loss": 0.7017,
+      "step": 4529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7685053143291274,
+      "learning_rate": 1.1746934990846013e-05,
+      "loss": 0.7724,
+      "step": 4530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.590333764801969,
+      "learning_rate": 1.1743691100013459e-05,
+      "loss": 0.7066,
+      "step": 4531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.556934087076597,
+      "learning_rate": 1.1740447019930506e-05,
+      "loss": 0.686,
+      "step": 4532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4977064506551216,
+      "learning_rate": 1.1737202750949244e-05,
+      "loss": 0.7134,
+      "step": 4533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5293106757310524,
+      "learning_rate": 1.173395829342179e-05,
+      "loss": 0.6573,
+      "step": 4534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4865247698778525,
+      "learning_rate": 1.1730713647700282e-05,
+      "loss": 0.6099,
+      "step": 4535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5621144654474877,
+      "learning_rate": 1.1727468814136871e-05,
+      "loss": 0.7107,
+      "step": 4536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9113889775095725,
+      "learning_rate": 1.1724223793083733e-05,
+      "loss": 0.81,
+      "step": 4537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3475550233522957,
+      "learning_rate": 1.172097858489307e-05,
+      "loss": 0.6461,
+      "step": 4538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6564171669585754,
+      "learning_rate": 1.1717733189917092e-05,
+      "loss": 0.722,
+      "step": 4539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6240145152483296,
+      "learning_rate": 1.1714487608508043e-05,
+      "loss": 0.699,
+      "step": 4540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7102134687250472,
+      "learning_rate": 1.1711241841018173e-05,
+      "loss": 0.6937,
+      "step": 4541
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6969571762175641,
+      "learning_rate": 1.1707995887799759e-05,
+      "loss": 0.7627,
+      "step": 4542
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7822040158464596,
+      "learning_rate": 1.1704749749205105e-05,
+      "loss": 0.8299,
+      "step": 4543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5846248801060048,
+      "learning_rate": 1.1701503425586522e-05,
+      "loss": 0.8152,
+      "step": 4544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.458804753241937,
+      "learning_rate": 1.1698256917296354e-05,
+      "loss": 0.7306,
+      "step": 4545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.509308899737421,
+      "learning_rate": 1.1695010224686951e-05,
+      "loss": 0.7077,
+      "step": 4546
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5690076410012654,
+      "learning_rate": 1.1691763348110698e-05,
+      "loss": 0.7536,
+      "step": 4547
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6779952948032286,
+      "learning_rate": 1.168851628791999e-05,
+      "loss": 0.7712,
+      "step": 4548
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7875031142218436,
+      "learning_rate": 1.1685269044467244e-05,
+      "loss": 0.8445,
+      "step": 4549
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4464393868159005,
+      "learning_rate": 1.16820216181049e-05,
+      "loss": 0.6798,
+      "step": 4550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5728041249582039,
+      "learning_rate": 1.167877400918541e-05,
+      "loss": 0.6674,
+      "step": 4551
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9005455827057687,
+      "learning_rate": 1.1675526218061254e-05,
+      "loss": 0.7243,
+      "step": 4552
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5418841385192381,
+      "learning_rate": 1.1672278245084931e-05,
+      "loss": 0.6592,
+      "step": 4553
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.495491500895745,
+      "learning_rate": 1.1669030090608957e-05,
+      "loss": 0.6856,
+      "step": 4554
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6104076342307279,
+      "learning_rate": 1.1665781754985867e-05,
+      "loss": 0.6973,
+      "step": 4555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5121867638957396,
+      "learning_rate": 1.1662533238568214e-05,
+      "loss": 0.7295,
+      "step": 4556
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5661736979533423,
+      "learning_rate": 1.1659284541708581e-05,
+      "loss": 0.7195,
+      "step": 4557
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.454697725959841,
+      "learning_rate": 1.165603566475956e-05,
+      "loss": 0.6999,
+      "step": 4558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4835668066546128,
+      "learning_rate": 1.1652786608073763e-05,
+      "loss": 0.7271,
+      "step": 4559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3864305015981893,
+      "learning_rate": 1.1649537372003825e-05,
+      "loss": 0.6977,
+      "step": 4560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5254952489151536,
+      "learning_rate": 1.1646287956902402e-05,
+      "loss": 0.6152,
+      "step": 4561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6481920827138161,
+      "learning_rate": 1.1643038363122168e-05,
+      "loss": 0.7651,
+      "step": 4562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4195624179129456,
+      "learning_rate": 1.1639788591015815e-05,
+      "loss": 0.6878,
+      "step": 4563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6036746756352342,
+      "learning_rate": 1.1636538640936052e-05,
+      "loss": 0.7982,
+      "step": 4564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5207564745319613,
+      "learning_rate": 1.163328851323561e-05,
+      "loss": 0.7535,
+      "step": 4565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5653334921119917,
+      "learning_rate": 1.1630038208267244e-05,
+      "loss": 0.7109,
+      "step": 4566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5478786441455192,
+      "learning_rate": 1.1626787726383722e-05,
+      "loss": 0.6924,
+      "step": 4567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5811167412057043,
+      "learning_rate": 1.1623537067937832e-05,
+      "loss": 0.766,
+      "step": 4568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5202293492541648,
+      "learning_rate": 1.162028623328238e-05,
+      "loss": 0.709,
+      "step": 4569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.538960478366557,
+      "learning_rate": 1.1617035222770199e-05,
+      "loss": 0.7687,
+      "step": 4570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4244896266657237,
+      "learning_rate": 1.1613784036754132e-05,
+      "loss": 0.6506,
+      "step": 4571
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5037212119134407,
+      "learning_rate": 1.1610532675587042e-05,
+      "loss": 0.6561,
+      "step": 4572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5084780381210663,
+      "learning_rate": 1.160728113962182e-05,
+      "loss": 0.6705,
+      "step": 4573
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7686645441932545,
+      "learning_rate": 1.1604029429211361e-05,
+      "loss": 0.7333,
+      "step": 4574
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4847152737520022,
+      "learning_rate": 1.1600777544708595e-05,
+      "loss": 0.7147,
+      "step": 4575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.550495765299229,
+      "learning_rate": 1.1597525486466458e-05,
+      "loss": 0.68,
+      "step": 4576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5757475793774263,
+      "learning_rate": 1.1594273254837914e-05,
+      "loss": 0.7538,
+      "step": 4577
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5793316860033737,
+      "learning_rate": 1.1591020850175935e-05,
+      "loss": 0.7194,
+      "step": 4578
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5312094094225086,
+      "learning_rate": 1.1587768272833529e-05,
+      "loss": 0.795,
+      "step": 4579
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6037209281036378,
+      "learning_rate": 1.1584515523163705e-05,
+      "loss": 0.6588,
+      "step": 4580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4918056952223486,
+      "learning_rate": 1.15812626015195e-05,
+      "loss": 0.6309,
+      "step": 4581
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.605269547265067,
+      "learning_rate": 1.1578009508253966e-05,
+      "loss": 0.7243,
+      "step": 4582
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4666143394088873,
+      "learning_rate": 1.157475624372018e-05,
+      "loss": 0.6769,
+      "step": 4583
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6405211541922342,
+      "learning_rate": 1.1571502808271229e-05,
+      "loss": 0.7228,
+      "step": 4584
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6232228814453615,
+      "learning_rate": 1.1568249202260224e-05,
+      "loss": 0.8883,
+      "step": 4585
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5683314189351627,
+      "learning_rate": 1.1564995426040293e-05,
+      "loss": 0.672,
+      "step": 4586
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.548145977674433,
+      "learning_rate": 1.1561741479964579e-05,
+      "loss": 0.6571,
+      "step": 4587
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6133717388005364,
+      "learning_rate": 1.1558487364386255e-05,
+      "loss": 0.7443,
+      "step": 4588
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5584829486616945,
+      "learning_rate": 1.1555233079658496e-05,
+      "loss": 0.7495,
+      "step": 4589
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5753794669064949,
+      "learning_rate": 1.155197862613451e-05,
+      "loss": 0.7855,
+      "step": 4590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4472244902213225,
+      "learning_rate": 1.1548724004167513e-05,
+      "loss": 0.7718,
+      "step": 4591
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5367327064558058,
+      "learning_rate": 1.1545469214110745e-05,
+      "loss": 0.8578,
+      "step": 4592
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5334893194368162,
+      "learning_rate": 1.1542214256317461e-05,
+      "loss": 0.83,
+      "step": 4593
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5910044241350005,
+      "learning_rate": 1.153895913114094e-05,
+      "loss": 0.6814,
+      "step": 4594
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6807368860123768,
+      "learning_rate": 1.153570383893447e-05,
+      "loss": 0.733,
+      "step": 4595
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5584840199996832,
+      "learning_rate": 1.1532448380051362e-05,
+      "loss": 0.6921,
+      "step": 4596
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5458643754720789,
+      "learning_rate": 1.152919275484495e-05,
+      "loss": 0.7061,
+      "step": 4597
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7702393765678885,
+      "learning_rate": 1.1525936963668579e-05,
+      "loss": 0.7618,
+      "step": 4598
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4375965231055472,
+      "learning_rate": 1.1522681006875614e-05,
+      "loss": 0.6703,
+      "step": 4599
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.697375977900443,
+      "learning_rate": 1.1519424884819432e-05,
+      "loss": 0.7234,
+      "step": 4600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4384659116884588,
+      "learning_rate": 1.1516168597853446e-05,
+      "loss": 0.7087,
+      "step": 4601
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.495867458271716,
+      "learning_rate": 1.1512912146331066e-05,
+      "loss": 0.6975,
+      "step": 4602
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5537088310355118,
+      "learning_rate": 1.1509655530605732e-05,
+      "loss": 0.8397,
+      "step": 4603
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.374743892023546,
+      "learning_rate": 1.1506398751030899e-05,
+      "loss": 0.6131,
+      "step": 4604
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.677575822578111,
+      "learning_rate": 1.150314180796004e-05,
+      "loss": 0.6829,
+      "step": 4605
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4445203506416073,
+      "learning_rate": 1.1499884701746642e-05,
+      "loss": 0.798,
+      "step": 4606
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.712464858100849,
+      "learning_rate": 1.1496627432744216e-05,
+      "loss": 0.7282,
+      "step": 4607
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5985773032288149,
+      "learning_rate": 1.1493370001306288e-05,
+      "loss": 0.7807,
+      "step": 4608
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.486347003648851,
+      "learning_rate": 1.14901124077864e-05,
+      "loss": 0.7325,
+      "step": 4609
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5134227711225614,
+      "learning_rate": 1.148685465253811e-05,
+      "loss": 0.6873,
+      "step": 4610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4965204314771694,
+      "learning_rate": 1.1483596735915003e-05,
+      "loss": 0.6983,
+      "step": 4611
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6605822099109793,
+      "learning_rate": 1.148033865827067e-05,
+      "loss": 0.6704,
+      "step": 4612
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2937035044904677,
+      "learning_rate": 1.1477080419958726e-05,
+      "loss": 0.6363,
+      "step": 4613
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8160972538634474,
+      "learning_rate": 1.14738220213328e-05,
+      "loss": 0.7374,
+      "step": 4614
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4711744385294088,
+      "learning_rate": 1.1470563462746542e-05,
+      "loss": 0.7394,
+      "step": 4615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8100258979621198,
+      "learning_rate": 1.1467304744553618e-05,
+      "loss": 0.7842,
+      "step": 4616
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4886306908892508,
+      "learning_rate": 1.1464045867107712e-05,
+      "loss": 0.7585,
+      "step": 4617
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4741574652866123,
+      "learning_rate": 1.1460786830762519e-05,
+      "loss": 0.6208,
+      "step": 4618
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5586042387688461,
+      "learning_rate": 1.1457527635871759e-05,
+      "loss": 0.6356,
+      "step": 4619
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6055025072357336,
+      "learning_rate": 1.145426828278917e-05,
+      "loss": 0.8323,
+      "step": 4620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6099059419948964,
+      "learning_rate": 1.14510087718685e-05,
+      "loss": 0.7571,
+      "step": 4621
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6203016711208587,
+      "learning_rate": 1.1447749103463519e-05,
+      "loss": 0.7162,
+      "step": 4622
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7047316755504125,
+      "learning_rate": 1.144448927792801e-05,
+      "loss": 0.8016,
+      "step": 4623
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5676650614066094,
+      "learning_rate": 1.1441229295615784e-05,
+      "loss": 0.7296,
+      "step": 4624
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.622912975230112,
+      "learning_rate": 1.1437969156880656e-05,
+      "loss": 0.764,
+      "step": 4625
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4111662466560349,
+      "learning_rate": 1.143470886207646e-05,
+      "loss": 0.6159,
+      "step": 4626
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5863421762962944,
+      "learning_rate": 1.1431448411557056e-05,
+      "loss": 0.7665,
+      "step": 4627
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6377625649613345,
+      "learning_rate": 1.1428187805676309e-05,
+      "loss": 0.7062,
+      "step": 4628
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4509682195293687,
+      "learning_rate": 1.1424927044788114e-05,
+      "loss": 0.6903,
+      "step": 4629
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5453389143146479,
+      "learning_rate": 1.1421666129246372e-05,
+      "loss": 0.7763,
+      "step": 4630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.483135906834292,
+      "learning_rate": 1.1418405059405002e-05,
+      "loss": 0.797,
+      "step": 4631
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7483319893470177,
+      "learning_rate": 1.1415143835617943e-05,
+      "loss": 0.8703,
+      "step": 4632
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5270226220355356,
+      "learning_rate": 1.1411882458239152e-05,
+      "loss": 0.7648,
+      "step": 4633
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6431099516940701,
+      "learning_rate": 1.14086209276226e-05,
+      "loss": 0.7689,
+      "step": 4634
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4282775689777298,
+      "learning_rate": 1.1405359244122275e-05,
+      "loss": 0.6884,
+      "step": 4635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6107562592794664,
+      "learning_rate": 1.1402097408092184e-05,
+      "loss": 0.7794,
+      "step": 4636
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7817765520103868,
+      "learning_rate": 1.1398835419886339e-05,
+      "loss": 0.7781,
+      "step": 4637
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7683257244565183,
+      "learning_rate": 1.1395573279858789e-05,
+      "loss": 0.7045,
+      "step": 4638
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6521730801992396,
+      "learning_rate": 1.1392310988363584e-05,
+      "loss": 0.75,
+      "step": 4639
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.521210187770528,
+      "learning_rate": 1.1389048545754794e-05,
+      "loss": 0.7056,
+      "step": 4640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5150310415725645,
+      "learning_rate": 1.1385785952386503e-05,
+      "loss": 0.7595,
+      "step": 4641
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3661564856473303,
+      "learning_rate": 1.1382523208612823e-05,
+      "loss": 0.7069,
+      "step": 4642
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6708812769448553,
+      "learning_rate": 1.1379260314787867e-05,
+      "loss": 0.7898,
+      "step": 4643
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6704991799723237,
+      "learning_rate": 1.1375997271265775e-05,
+      "loss": 0.77,
+      "step": 4644
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5237140033724268,
+      "learning_rate": 1.1372734078400695e-05,
+      "loss": 0.7653,
+      "step": 4645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5858634760779247,
+      "learning_rate": 1.1369470736546798e-05,
+      "loss": 0.8662,
+      "step": 4646
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4565672294570617,
+      "learning_rate": 1.1366207246058269e-05,
+      "loss": 0.7507,
+      "step": 4647
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4991843578593989,
+      "learning_rate": 1.136294360728931e-05,
+      "loss": 0.7551,
+      "step": 4648
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5417235142162893,
+      "learning_rate": 1.1359679820594137e-05,
+      "loss": 0.6552,
+      "step": 4649
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.548589989786535,
+      "learning_rate": 1.135641588632698e-05,
+      "loss": 0.611,
+      "step": 4650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4696288433018039,
+      "learning_rate": 1.1353151804842088e-05,
+      "loss": 0.7258,
+      "step": 4651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4592926096276786,
+      "learning_rate": 1.1349887576493735e-05,
+      "loss": 0.6607,
+      "step": 4652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6058655036855474,
+      "learning_rate": 1.1346623201636192e-05,
+      "loss": 0.7338,
+      "step": 4653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.580983795741087,
+      "learning_rate": 1.1343358680623757e-05,
+      "loss": 0.8141,
+      "step": 4654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3749699175665049,
+      "learning_rate": 1.1340094013810749e-05,
+      "loss": 0.6731,
+      "step": 4655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6096032470138852,
+      "learning_rate": 1.1336829201551492e-05,
+      "loss": 0.7535,
+      "step": 4656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6262893926298707,
+      "learning_rate": 1.1333564244200328e-05,
+      "loss": 0.7377,
+      "step": 4657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5410648418608643,
+      "learning_rate": 1.1330299142111623e-05,
+      "loss": 0.739,
+      "step": 4658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5485513987128794,
+      "learning_rate": 1.1327033895639747e-05,
+      "loss": 0.8029,
+      "step": 4659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.410059273776502,
+      "learning_rate": 1.1323768505139095e-05,
+      "loss": 0.7232,
+      "step": 4660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5624646887742166,
+      "learning_rate": 1.1320502970964076e-05,
+      "loss": 0.7231,
+      "step": 4661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5372926149461579,
+      "learning_rate": 1.1317237293469108e-05,
+      "loss": 0.7501,
+      "step": 4662
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5681818631952515,
+      "learning_rate": 1.1313971473008628e-05,
+      "loss": 0.7218,
+      "step": 4663
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.482830029646937,
+      "learning_rate": 1.1310705509937096e-05,
+      "loss": 0.7101,
+      "step": 4664
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.562764113175667,
+      "learning_rate": 1.1307439404608978e-05,
+      "loss": 0.7555,
+      "step": 4665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4911542811063758,
+      "learning_rate": 1.130417315737876e-05,
+      "loss": 0.6715,
+      "step": 4666
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5254406078226133,
+      "learning_rate": 1.1300906768600939e-05,
+      "loss": 0.6447,
+      "step": 4667
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4471085397289045,
+      "learning_rate": 1.129764023863003e-05,
+      "loss": 0.7193,
+      "step": 4668
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5608343873438544,
+      "learning_rate": 1.1294373567820568e-05,
+      "loss": 0.661,
+      "step": 4669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4437985820957768,
+      "learning_rate": 1.12911067565271e-05,
+      "loss": 0.624,
+      "step": 4670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4246038270133796,
+      "learning_rate": 1.1287839805104182e-05,
+      "loss": 0.691,
+      "step": 4671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4721353611911032,
+      "learning_rate": 1.1284572713906392e-05,
+      "loss": 0.7062,
+      "step": 4672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7658949422577648,
+      "learning_rate": 1.1281305483288321e-05,
+      "loss": 0.7626,
+      "step": 4673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.504533207614064,
+      "learning_rate": 1.1278038113604579e-05,
+      "loss": 0.7198,
+      "step": 4674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5931060163297304,
+      "learning_rate": 1.1274770605209786e-05,
+      "loss": 0.8055,
+      "step": 4675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5542388836057903,
+      "learning_rate": 1.1271502958458577e-05,
+      "loss": 0.7602,
+      "step": 4676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.575193549763372,
+      "learning_rate": 1.1268235173705608e-05,
+      "loss": 0.8088,
+      "step": 4677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4411760933441318,
+      "learning_rate": 1.1264967251305542e-05,
+      "loss": 0.6863,
+      "step": 4678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6648858012128047,
+      "learning_rate": 1.1261699191613067e-05,
+      "loss": 0.7772,
+      "step": 4679
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7199672360791733,
+      "learning_rate": 1.1258430994982872e-05,
+      "loss": 0.8195,
+      "step": 4680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4568557243100269,
+      "learning_rate": 1.1255162661769674e-05,
+      "loss": 0.6591,
+      "step": 4681
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.520795016619549,
+      "learning_rate": 1.1251894192328194e-05,
+      "loss": 0.7553,
+      "step": 4682
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6015181193109884,
+      "learning_rate": 1.124862558701318e-05,
+      "loss": 0.7741,
+      "step": 4683
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4447539094155426,
+      "learning_rate": 1.1245356846179384e-05,
+      "loss": 0.7089,
+      "step": 4684
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4584951955485221,
+      "learning_rate": 1.1242087970181578e-05,
+      "loss": 0.7395,
+      "step": 4685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4461603196919004,
+      "learning_rate": 1.1238818959374547e-05,
+      "loss": 0.6715,
+      "step": 4686
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7311798026421528,
+      "learning_rate": 1.1235549814113092e-05,
+      "loss": 0.8515,
+      "step": 4687
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4048144176782515,
+      "learning_rate": 1.1232280534752025e-05,
+      "loss": 0.5922,
+      "step": 4688
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5481973447074155,
+      "learning_rate": 1.1229011121646176e-05,
+      "loss": 0.7715,
+      "step": 4689
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6958381976979935,
+      "learning_rate": 1.1225741575150391e-05,
+      "loss": 0.7143,
+      "step": 4690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4660906832812384,
+      "learning_rate": 1.1222471895619525e-05,
+      "loss": 0.7651,
+      "step": 4691
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.514977911309017,
+      "learning_rate": 1.1219202083408454e-05,
+      "loss": 0.7018,
+      "step": 4692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7532823984411445,
+      "learning_rate": 1.1215932138872061e-05,
+      "loss": 0.7323,
+      "step": 4693
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4676790184219108,
+      "learning_rate": 1.1212662062365252e-05,
+      "loss": 0.7031,
+      "step": 4694
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6054557795695394,
+      "learning_rate": 1.1209391854242938e-05,
+      "loss": 0.8304,
+      "step": 4695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5991701867348644,
+      "learning_rate": 1.120612151486005e-05,
+      "loss": 0.6987,
+      "step": 4696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5170179933755257,
+      "learning_rate": 1.1202851044571533e-05,
+      "loss": 0.6792,
+      "step": 4697
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7157827956741485,
+      "learning_rate": 1.1199580443732347e-05,
+      "loss": 0.7554,
+      "step": 4698
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4753493754003022,
+      "learning_rate": 1.1196309712697463e-05,
+      "loss": 0.6668,
+      "step": 4699
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4924475104111545,
+      "learning_rate": 1.1193038851821867e-05,
+      "loss": 0.7118,
+      "step": 4700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5430843931734841,
+      "learning_rate": 1.1189767861460563e-05,
+      "loss": 0.7068,
+      "step": 4701
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5076245973794509,
+      "learning_rate": 1.1186496741968562e-05,
+      "loss": 0.6831,
+      "step": 4702
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7370968200999768,
+      "learning_rate": 1.1183225493700895e-05,
+      "loss": 0.7394,
+      "step": 4703
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.800654570183203,
+      "learning_rate": 1.1179954117012603e-05,
+      "loss": 0.7211,
+      "step": 4704
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5017002098978107,
+      "learning_rate": 1.1176682612258745e-05,
+      "loss": 0.7532,
+      "step": 4705
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6617996193529947,
+      "learning_rate": 1.117341097979439e-05,
+      "loss": 0.7171,
+      "step": 4706
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7063874397456549,
+      "learning_rate": 1.1170139219974626e-05,
+      "loss": 0.7891,
+      "step": 4707
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6636920434588713,
+      "learning_rate": 1.1166867333154543e-05,
+      "loss": 0.7299,
+      "step": 4708
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5654818487601652,
+      "learning_rate": 1.1163595319689264e-05,
+      "loss": 0.6421,
+      "step": 4709
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4218850872715825,
+      "learning_rate": 1.1160323179933908e-05,
+      "loss": 0.6249,
+      "step": 4710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5576364310269057,
+      "learning_rate": 1.1157050914243614e-05,
+      "loss": 0.7484,
+      "step": 4711
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6310208232754702,
+      "learning_rate": 1.115377852297354e-05,
+      "loss": 0.7097,
+      "step": 4712
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6373612010247085,
+      "learning_rate": 1.1150506006478849e-05,
+      "loss": 0.7567,
+      "step": 4713
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.764893872771686,
+      "learning_rate": 1.1147233365114725e-05,
+      "loss": 0.7486,
+      "step": 4714
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4628182364711966,
+      "learning_rate": 1.1143960599236358e-05,
+      "loss": 0.6277,
+      "step": 4715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7369395077416692,
+      "learning_rate": 1.1140687709198958e-05,
+      "loss": 0.7288,
+      "step": 4716
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.635015653974785,
+      "learning_rate": 1.1137414695357747e-05,
+      "loss": 0.7613,
+      "step": 4717
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5503660871383618,
+      "learning_rate": 1.1134141558067955e-05,
+      "loss": 0.7288,
+      "step": 4718
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5538599617254092,
+      "learning_rate": 1.1130868297684833e-05,
+      "loss": 0.7645,
+      "step": 4719
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.511799580256174,
+      "learning_rate": 1.1127594914563647e-05,
+      "loss": 0.8259,
+      "step": 4720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.557814325422162,
+      "learning_rate": 1.1124321409059661e-05,
+      "loss": 0.7098,
+      "step": 4721
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5253434446450445,
+      "learning_rate": 1.1121047781528174e-05,
+      "loss": 0.7664,
+      "step": 4722
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7265330902888985,
+      "learning_rate": 1.1117774032324476e-05,
+      "loss": 0.7295,
+      "step": 4723
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5518678435153568,
+      "learning_rate": 1.111450016180389e-05,
+      "loss": 0.7724,
+      "step": 4724
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5664907417938099,
+      "learning_rate": 1.1111226170321738e-05,
+      "loss": 0.7906,
+      "step": 4725
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5130135727104113,
+      "learning_rate": 1.1107952058233367e-05,
+      "loss": 0.766,
+      "step": 4726
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2548589391419385,
+      "learning_rate": 1.1104677825894121e-05,
+      "loss": 0.7856,
+      "step": 4727
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4289814666356178,
+      "learning_rate": 1.1101403473659376e-05,
+      "loss": 0.7476,
+      "step": 4728
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6747946340533135,
+      "learning_rate": 1.1098129001884508e-05,
+      "loss": 0.8052,
+      "step": 4729
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5331014347391687,
+      "learning_rate": 1.1094854410924909e-05,
+      "loss": 0.8508,
+      "step": 4730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8579945737436387,
+      "learning_rate": 1.1091579701135986e-05,
+      "loss": 0.7624,
+      "step": 4731
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.566768370211104,
+      "learning_rate": 1.1088304872873152e-05,
+      "loss": 0.7611,
+      "step": 4732
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9112271223739234,
+      "learning_rate": 1.1085029926491848e-05,
+      "loss": 0.8016,
+      "step": 4733
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.507987571004608,
+      "learning_rate": 1.1081754862347513e-05,
+      "loss": 0.7083,
+      "step": 4734
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4228665977224153,
+      "learning_rate": 1.1078479680795604e-05,
+      "loss": 0.7893,
+      "step": 4735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6040397098394448,
+      "learning_rate": 1.107520438219159e-05,
+      "loss": 0.7328,
+      "step": 4736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5392882633332718,
+      "learning_rate": 1.1071928966890957e-05,
+      "loss": 0.781,
+      "step": 4737
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.666983043219811,
+      "learning_rate": 1.1068653435249197e-05,
+      "loss": 0.7516,
+      "step": 4738
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6269538017114316,
+      "learning_rate": 1.1065377787621819e-05,
+      "loss": 0.7179,
+      "step": 4739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.714058517254383,
+      "learning_rate": 1.1062102024364341e-05,
+      "loss": 0.786,
+      "step": 4740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6659851540381478,
+      "learning_rate": 1.1058826145832298e-05,
+      "loss": 0.8467,
+      "step": 4741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6685789483122337,
+      "learning_rate": 1.1055550152381235e-05,
+      "loss": 0.7775,
+      "step": 4742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6167305758786248,
+      "learning_rate": 1.1052274044366711e-05,
+      "loss": 0.7274,
+      "step": 4743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4746836673435093,
+      "learning_rate": 1.1048997822144296e-05,
+      "loss": 0.8207,
+      "step": 4744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.633586522725332,
+      "learning_rate": 1.1045721486069568e-05,
+      "loss": 0.7326,
+      "step": 4745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4809266073703506,
+      "learning_rate": 1.1042445036498129e-05,
+      "loss": 0.7137,
+      "step": 4746
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5195680421319833,
+      "learning_rate": 1.1039168473785584e-05,
+      "loss": 0.7911,
+      "step": 4747
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6137137131166324,
+      "learning_rate": 1.103589179828755e-05,
+      "loss": 0.7207,
+      "step": 4748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4873281881742182,
+      "learning_rate": 1.1032615010359661e-05,
+      "loss": 0.672,
+      "step": 4749
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5131039557796713,
+      "learning_rate": 1.1029338110357561e-05,
+      "loss": 0.5865,
+      "step": 4750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4201459792639004,
+      "learning_rate": 1.1026061098636906e-05,
+      "loss": 0.6513,
+      "step": 4751
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.486349912355769,
+      "learning_rate": 1.1022783975553366e-05,
+      "loss": 0.6578,
+      "step": 4752
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5176489180640584,
+      "learning_rate": 1.101950674146262e-05,
+      "loss": 0.6687,
+      "step": 4753
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6306745452948275,
+      "learning_rate": 1.101622939672036e-05,
+      "loss": 0.6677,
+      "step": 4754
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.78403339360376,
+      "learning_rate": 1.1012951941682291e-05,
+      "loss": 0.747,
+      "step": 4755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5074978240557007,
+      "learning_rate": 1.1009674376704128e-05,
+      "loss": 0.751,
+      "step": 4756
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.539491194858523,
+      "learning_rate": 1.1006396702141605e-05,
+      "loss": 0.7455,
+      "step": 4757
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7450620289002257,
+      "learning_rate": 1.1003118918350456e-05,
+      "loss": 0.7322,
+      "step": 4758
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5733533171495808,
+      "learning_rate": 1.0999841025686431e-05,
+      "loss": 0.5972,
+      "step": 4759
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5719877764421808,
+      "learning_rate": 1.0996563024505303e-05,
+      "loss": 0.7433,
+      "step": 4760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3363325039311442,
+      "learning_rate": 1.0993284915162843e-05,
+      "loss": 0.6251,
+      "step": 4761
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4101295343750289,
+      "learning_rate": 1.0990006698014837e-05,
+      "loss": 0.6571,
+      "step": 4762
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3324377019912343,
+      "learning_rate": 1.0986728373417088e-05,
+      "loss": 0.5947,
+      "step": 4763
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6333620766873393,
+      "learning_rate": 1.0983449941725404e-05,
+      "loss": 0.7579,
+      "step": 4764
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6989820977183576,
+      "learning_rate": 1.098017140329561e-05,
+      "loss": 0.6614,
+      "step": 4765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5800683448962918,
+      "learning_rate": 1.0976892758483533e-05,
+      "loss": 0.6856,
+      "step": 4766
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5916909083774833,
+      "learning_rate": 1.0973614007645027e-05,
+      "loss": 0.7061,
+      "step": 4767
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4914111133780676,
+      "learning_rate": 1.097033515113595e-05,
+      "loss": 0.6408,
+      "step": 4768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.657441182352292,
+      "learning_rate": 1.096705618931216e-05,
+      "loss": 0.6737,
+      "step": 4769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6302751638129835,
+      "learning_rate": 1.0963777122529547e-05,
+      "loss": 0.6715,
+      "step": 4770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5698054269632449,
+      "learning_rate": 1.0960497951144001e-05,
+      "loss": 0.8629,
+      "step": 4771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5812627655597422,
+      "learning_rate": 1.0957218675511418e-05,
+      "loss": 0.7924,
+      "step": 4772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6161084006625803,
+      "learning_rate": 1.0953939295987722e-05,
+      "loss": 0.7627,
+      "step": 4773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6299841409594045,
+      "learning_rate": 1.095065981292883e-05,
+      "loss": 0.6798,
+      "step": 4774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.588289422502915,
+      "learning_rate": 1.0947380226690686e-05,
+      "loss": 0.6632,
+      "step": 4775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7083918017260773,
+      "learning_rate": 1.0944100537629229e-05,
+      "loss": 0.7163,
+      "step": 4776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4988568694539077,
+      "learning_rate": 1.0940820746100425e-05,
+      "loss": 0.758,
+      "step": 4777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4764122582404338,
+      "learning_rate": 1.0937540852460242e-05,
+      "loss": 0.7082,
+      "step": 4778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4720694527200804,
+      "learning_rate": 1.093426085706466e-05,
+      "loss": 0.7218,
+      "step": 4779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5462059060882711,
+      "learning_rate": 1.0930980760269673e-05,
+      "loss": 0.7314,
+      "step": 4780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4346137380421373,
+      "learning_rate": 1.0927700562431283e-05,
+      "loss": 0.695,
+      "step": 4781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5832976035826196,
+      "learning_rate": 1.0924420263905505e-05,
+      "loss": 0.7294,
+      "step": 4782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.557316426835322,
+      "learning_rate": 1.0921139865048362e-05,
+      "loss": 0.7435,
+      "step": 4783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4448128642118385,
+      "learning_rate": 1.0917859366215895e-05,
+      "loss": 0.7079,
+      "step": 4784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5853405092750676,
+      "learning_rate": 1.0914578767764145e-05,
+      "loss": 0.7592,
+      "step": 4785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4668078137037874,
+      "learning_rate": 1.0911298070049172e-05,
+      "loss": 0.7091,
+      "step": 4786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4979091711758297,
+      "learning_rate": 1.0908017273427044e-05,
+      "loss": 0.7898,
+      "step": 4787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4492199846891123,
+      "learning_rate": 1.0904736378253844e-05,
+      "loss": 0.6762,
+      "step": 4788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4721752479841992,
+      "learning_rate": 1.0901455384885658e-05,
+      "loss": 0.6579,
+      "step": 4789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.428659963881077,
+      "learning_rate": 1.0898174293678583e-05,
+      "loss": 0.6705,
+      "step": 4790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6378135027870262,
+      "learning_rate": 1.0894893104988738e-05,
+      "loss": 0.7132,
+      "step": 4791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.556497741824382,
+      "learning_rate": 1.089161181917224e-05,
+      "loss": 0.7206,
+      "step": 4792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7230568439952965,
+      "learning_rate": 1.0888330436585224e-05,
+      "loss": 0.6613,
+      "step": 4793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.612851513748318,
+      "learning_rate": 1.088504895758383e-05,
+      "loss": 0.7709,
+      "step": 4794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4659086390595735,
+      "learning_rate": 1.088176738252421e-05,
+      "loss": 0.6839,
+      "step": 4795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9422400062295213,
+      "learning_rate": 1.0878485711762533e-05,
+      "loss": 0.8199,
+      "step": 4796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5666072973905,
+      "learning_rate": 1.0875203945654969e-05,
+      "loss": 0.758,
+      "step": 4797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4006583596790678,
+      "learning_rate": 1.0871922084557702e-05,
+      "loss": 0.6683,
+      "step": 4798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5592372404936563,
+      "learning_rate": 1.0868640128826929e-05,
+      "loss": 0.7184,
+      "step": 4799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.392316596997383,
+      "learning_rate": 1.0865358078818855e-05,
+      "loss": 0.6242,
+      "step": 4800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4557220760954837,
+      "learning_rate": 1.0862075934889694e-05,
+      "loss": 0.6217,
+      "step": 4801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6556912744299246,
+      "learning_rate": 1.085879369739567e-05,
+      "loss": 0.6765,
+      "step": 4802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5322411610345907,
+      "learning_rate": 1.085551136669302e-05,
+      "loss": 0.6711,
+      "step": 4803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5823221318504805,
+      "learning_rate": 1.0852228943137992e-05,
+      "loss": 0.703,
+      "step": 4804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5657096226939102,
+      "learning_rate": 1.0848946427086839e-05,
+      "loss": 0.7833,
+      "step": 4805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.452125827239233,
+      "learning_rate": 1.0845663818895826e-05,
+      "loss": 0.5953,
+      "step": 4806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4984899358281312,
+      "learning_rate": 1.0842381118921233e-05,
+      "loss": 0.6292,
+      "step": 4807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.705995215029523,
+      "learning_rate": 1.083909832751934e-05,
+      "loss": 0.7315,
+      "step": 4808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6352719630144974,
+      "learning_rate": 1.0835815445046447e-05,
+      "loss": 0.7375,
+      "step": 4809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6271663906870726,
+      "learning_rate": 1.0832532471858857e-05,
+      "loss": 0.8057,
+      "step": 4810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4259725059873003,
+      "learning_rate": 1.0829249408312888e-05,
+      "loss": 0.6011,
+      "step": 4811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5507813456773818,
+      "learning_rate": 1.0825966254764864e-05,
+      "loss": 0.7963,
+      "step": 4812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4161423181330814,
+      "learning_rate": 1.0822683011571118e-05,
+      "loss": 0.7603,
+      "step": 4813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7237254932378825,
+      "learning_rate": 1.0819399679087997e-05,
+      "loss": 0.7259,
+      "step": 4814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.672925153363037,
+      "learning_rate": 1.0816116257671856e-05,
+      "loss": 0.7901,
+      "step": 4815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6042468592684607,
+      "learning_rate": 1.0812832747679054e-05,
+      "loss": 0.6788,
+      "step": 4816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.496471070682988,
+      "learning_rate": 1.0809549149465971e-05,
+      "loss": 0.7245,
+      "step": 4817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6801939660509613,
+      "learning_rate": 1.0806265463388989e-05,
+      "loss": 0.6754,
+      "step": 4818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7168952904835155,
+      "learning_rate": 1.0802981689804499e-05,
+      "loss": 0.7401,
+      "step": 4819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5862059151970802,
+      "learning_rate": 1.0799697829068903e-05,
+      "loss": 0.8646,
+      "step": 4820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.504679180024265,
+      "learning_rate": 1.0796413881538614e-05,
+      "loss": 0.7113,
+      "step": 4821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5559146837534314,
+      "learning_rate": 1.0793129847570052e-05,
+      "loss": 0.6912,
+      "step": 4822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4792854093478405,
+      "learning_rate": 1.0789845727519647e-05,
+      "loss": 0.6655,
+      "step": 4823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3621064320761462,
+      "learning_rate": 1.0786561521743843e-05,
+      "loss": 0.7083,
+      "step": 4824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5724233675494086,
+      "learning_rate": 1.0783277230599088e-05,
+      "loss": 0.6914,
+      "step": 4825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6449096777501475,
+      "learning_rate": 1.0779992854441833e-05,
+      "loss": 0.7233,
+      "step": 4826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5395166521358816,
+      "learning_rate": 1.0776708393628555e-05,
+      "loss": 0.7435,
+      "step": 4827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5538379810180913,
+      "learning_rate": 1.0773423848515728e-05,
+      "loss": 0.6461,
+      "step": 4828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7966878586338726,
+      "learning_rate": 1.0770139219459834e-05,
+      "loss": 0.7655,
+      "step": 4829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6554512917828683,
+      "learning_rate": 1.0766854506817374e-05,
+      "loss": 0.7496,
+      "step": 4830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5582376751493443,
+      "learning_rate": 1.0763569710944848e-05,
+      "loss": 0.7463,
+      "step": 4831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9045453022232466,
+      "learning_rate": 1.076028483219877e-05,
+      "loss": 0.7948,
+      "step": 4832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6515591767481628,
+      "learning_rate": 1.0756999870935668e-05,
+      "loss": 0.8395,
+      "step": 4833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4945025057050394,
+      "learning_rate": 1.0753714827512063e-05,
+      "loss": 0.7287,
+      "step": 4834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.586248043237468,
+      "learning_rate": 1.0750429702284499e-05,
+      "loss": 0.7354,
+      "step": 4835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.567076394776895,
+      "learning_rate": 1.0747144495609531e-05,
+      "loss": 0.7669,
+      "step": 4836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6765578455669068,
+      "learning_rate": 1.074385920784371e-05,
+      "loss": 0.7558,
+      "step": 4837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.512168260010115,
+      "learning_rate": 1.0740573839343607e-05,
+      "loss": 0.829,
+      "step": 4838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4896477079673143,
+      "learning_rate": 1.0737288390465792e-05,
+      "loss": 0.6725,
+      "step": 4839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4702480151410497,
+      "learning_rate": 1.073400286156685e-05,
+      "loss": 0.7041,
+      "step": 4840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6569811631390055,
+      "learning_rate": 1.073071725300338e-05,
+      "loss": 0.7568,
+      "step": 4841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6490024318489755,
+      "learning_rate": 1.0727431565131978e-05,
+      "loss": 0.7976,
+      "step": 4842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5752260202004427,
+      "learning_rate": 1.0724145798309258e-05,
+      "loss": 0.714,
+      "step": 4843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5449459528099991,
+      "learning_rate": 1.0720859952891834e-05,
+      "loss": 0.7015,
+      "step": 4844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6503976761089003,
+      "learning_rate": 1.0717574029236333e-05,
+      "loss": 0.8448,
+      "step": 4845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.612022678078976,
+      "learning_rate": 1.0714288027699392e-05,
+      "loss": 0.7325,
+      "step": 4846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6225715275901382,
+      "learning_rate": 1.071100194863766e-05,
+      "loss": 0.7276,
+      "step": 4847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.476340849753411,
+      "learning_rate": 1.0707715792407784e-05,
+      "loss": 0.726,
+      "step": 4848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4837181460190332,
+      "learning_rate": 1.0704429559366424e-05,
+      "loss": 0.6483,
+      "step": 4849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5986402635920125,
+      "learning_rate": 1.0701143249870253e-05,
+      "loss": 0.6338,
+      "step": 4850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7729055343245756,
+      "learning_rate": 1.0697856864275949e-05,
+      "loss": 0.8581,
+      "step": 4851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4999791982867494,
+      "learning_rate": 1.0694570402940192e-05,
+      "loss": 0.756,
+      "step": 4852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4845152414841658,
+      "learning_rate": 1.0691283866219682e-05,
+      "loss": 0.7603,
+      "step": 4853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5932376783084876,
+      "learning_rate": 1.0687997254471117e-05,
+      "loss": 0.6744,
+      "step": 4854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.685232234665782,
+      "learning_rate": 1.0684710568051211e-05,
+      "loss": 0.734,
+      "step": 4855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7060391227861584,
+      "learning_rate": 1.068142380731668e-05,
+      "loss": 0.764,
+      "step": 4856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.53478915495137,
+      "learning_rate": 1.0678136972624249e-05,
+      "loss": 0.7804,
+      "step": 4857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5942102451563043,
+      "learning_rate": 1.0674850064330655e-05,
+      "loss": 0.7312,
+      "step": 4858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5305582484928222,
+      "learning_rate": 1.0671563082792643e-05,
+      "loss": 0.624,
+      "step": 4859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7287157430425841,
+      "learning_rate": 1.0668276028366957e-05,
+      "loss": 0.7495,
+      "step": 4860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5015860213975631,
+      "learning_rate": 1.0664988901410361e-05,
+      "loss": 0.6169,
+      "step": 4861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5295651950629567,
+      "learning_rate": 1.0661701702279619e-05,
+      "loss": 0.7288,
+      "step": 4862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3833428272326505,
+      "learning_rate": 1.0658414431331502e-05,
+      "loss": 0.7572,
+      "step": 4863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8452863333294058,
+      "learning_rate": 1.0655127088922799e-05,
+      "loss": 0.7759,
+      "step": 4864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5262039228128124,
+      "learning_rate": 1.0651839675410295e-05,
+      "loss": 0.7951,
+      "step": 4865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5133758309522023,
+      "learning_rate": 1.0648552191150784e-05,
+      "loss": 0.5965,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.473671675400503,
+      "learning_rate": 1.0645264636501078e-05,
+      "loss": 0.7426,
+      "step": 4867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5734654354890787,
+      "learning_rate": 1.0641977011817986e-05,
+      "loss": 0.7293,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5238583307710616,
+      "learning_rate": 1.0638689317458328e-05,
+      "loss": 0.667,
+      "step": 4869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5674873789425725,
+      "learning_rate": 1.0635401553778934e-05,
+      "loss": 0.7048,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.476556899148976,
+      "learning_rate": 1.0632113721136636e-05,
+      "loss": 0.693,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5318947628197657,
+      "learning_rate": 1.0628825819888281e-05,
+      "loss": 0.7025,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.317050099437974,
+      "learning_rate": 1.0625537850390718e-05,
+      "loss": 0.7165,
+      "step": 4873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.591593406876283,
+      "learning_rate": 1.0622249813000799e-05,
+      "loss": 0.6269,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.521353505316856,
+      "learning_rate": 1.0618961708075398e-05,
+      "loss": 0.6831,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6774101897798614,
+      "learning_rate": 1.061567353597138e-05,
+      "loss": 0.8003,
+      "step": 4876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5712174893359296,
+      "learning_rate": 1.061238529704563e-05,
+      "loss": 0.7023,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5083574288941524,
+      "learning_rate": 1.060909699165503e-05,
+      "loss": 0.6789,
+      "step": 4878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7912885880405833,
+      "learning_rate": 1.060580862015648e-05,
+      "loss": 0.7922,
+      "step": 4879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4282792013974326,
+      "learning_rate": 1.0602520182906877e-05,
+      "loss": 0.6505,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.579379700276147,
+      "learning_rate": 1.0599231680263127e-05,
+      "loss": 0.688,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4453470872146301,
+      "learning_rate": 1.0595943112582152e-05,
+      "loss": 0.6834,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5607007989926776,
+      "learning_rate": 1.0592654480220874e-05,
+      "loss": 0.5767,
+      "step": 4883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5797955334370553,
+      "learning_rate": 1.0589365783536218e-05,
+      "loss": 0.6384,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5695344367475006,
+      "learning_rate": 1.0586077022885122e-05,
+      "loss": 0.7588,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6856965266457415,
+      "learning_rate": 1.0582788198624532e-05,
+      "loss": 0.7736,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5489583302375258,
+      "learning_rate": 1.0579499311111397e-05,
+      "loss": 0.7299,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5112208970054677,
+      "learning_rate": 1.0576210360702673e-05,
+      "loss": 0.7585,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.630106670056698,
+      "learning_rate": 1.0572921347755325e-05,
+      "loss": 0.6544,
+      "step": 4889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4713329795106127,
+      "learning_rate": 1.0569632272626323e-05,
+      "loss": 0.677,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5021811281953732,
+      "learning_rate": 1.0566343135672647e-05,
+      "loss": 0.7436,
+      "step": 4891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6665198712378295,
+      "learning_rate": 1.0563053937251282e-05,
+      "loss": 0.7593,
+      "step": 4892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5413742120634135,
+      "learning_rate": 1.0559764677719218e-05,
+      "loss": 0.731,
+      "step": 4893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.694579223239616,
+      "learning_rate": 1.0556475357433446e-05,
+      "loss": 0.6751,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.66881771693106,
+      "learning_rate": 1.0553185976750981e-05,
+      "loss": 0.7052,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.570220760256492,
+      "learning_rate": 1.054989653602883e-05,
+      "loss": 0.633,
+      "step": 4896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6084551011215464,
+      "learning_rate": 1.0546607035624008e-05,
+      "loss": 0.842,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5322621024852836,
+      "learning_rate": 1.0543317475893541e-05,
+      "loss": 0.7698,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6600600782491488,
+      "learning_rate": 1.0540027857194457e-05,
+      "loss": 0.7044,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.381638149166171,
+      "learning_rate": 1.0536738179883795e-05,
+      "loss": 0.6612,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.419555536907255,
+      "learning_rate": 1.0533448444318598e-05,
+      "loss": 0.6967,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6555120053495183,
+      "learning_rate": 1.0530158650855913e-05,
+      "loss": 0.7314,
+      "step": 4902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6046414041018124,
+      "learning_rate": 1.0526868799852797e-05,
+      "loss": 0.744,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7339299401857255,
+      "learning_rate": 1.0523578891666313e-05,
+      "loss": 0.7227,
+      "step": 4904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5164595375565013,
+      "learning_rate": 1.0520288926653528e-05,
+      "loss": 0.7927,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5925848482022713,
+      "learning_rate": 1.0516998905171516e-05,
+      "loss": 0.6613,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.614041586169326,
+      "learning_rate": 1.051370882757736e-05,
+      "loss": 0.7167,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.690770155753182,
+      "learning_rate": 1.0510418694228139e-05,
+      "loss": 0.7274,
+      "step": 4908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.558572494279657,
+      "learning_rate": 1.0507128505480956e-05,
+      "loss": 0.6946,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.64031812904918,
+      "learning_rate": 1.0503838261692902e-05,
+      "loss": 0.7701,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.461769504781515,
+      "learning_rate": 1.0500547963221086e-05,
+      "loss": 0.6864,
+      "step": 4911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5102708091137018,
+      "learning_rate": 1.0497257610422614e-05,
+      "loss": 0.6578,
+      "step": 4912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.863032025828687,
+      "learning_rate": 1.0493967203654604e-05,
+      "loss": 0.682,
+      "step": 4913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.383540654360992,
+      "learning_rate": 1.0490676743274181e-05,
+      "loss": 0.7154,
+      "step": 4914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5303593280040118,
+      "learning_rate": 1.0487386229638473e-05,
+      "loss": 0.7482,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4621120752135677,
+      "learning_rate": 1.0484095663104608e-05,
+      "loss": 0.6405,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.604206772487237,
+      "learning_rate": 1.048080504402973e-05,
+      "loss": 0.6942,
+      "step": 4917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6209202519114927,
+      "learning_rate": 1.0477514372770983e-05,
+      "loss": 0.7303,
+      "step": 4918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5004633451450609,
+      "learning_rate": 1.0474223649685517e-05,
+      "loss": 0.67,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6091999763739429,
+      "learning_rate": 1.0470932875130493e-05,
+      "loss": 0.6157,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6686612577304816,
+      "learning_rate": 1.0467642049463068e-05,
+      "loss": 0.7565,
+      "step": 4921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6312257099474892,
+      "learning_rate": 1.0464351173040412e-05,
+      "loss": 0.7045,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7194543341860458,
+      "learning_rate": 1.0461060246219699e-05,
+      "loss": 0.7014,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.477713446193529,
+      "learning_rate": 1.0457769269358104e-05,
+      "loss": 0.661,
+      "step": 4924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5176551779973013,
+      "learning_rate": 1.0454478242812814e-05,
+      "loss": 0.7093,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6337582845395682,
+      "learning_rate": 1.0451187166941019e-05,
+      "loss": 0.7892,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5818579907899044,
+      "learning_rate": 1.0447896042099914e-05,
+      "loss": 0.7644,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5478429707052999,
+      "learning_rate": 1.0444604868646695e-05,
+      "loss": 0.8387,
+      "step": 4928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5116126932431422,
+      "learning_rate": 1.0441313646938571e-05,
+      "loss": 0.7394,
+      "step": 4929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3437106808936985,
+      "learning_rate": 1.0438022377332755e-05,
+      "loss": 0.5898,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5724323804816993,
+      "learning_rate": 1.0434731060186456e-05,
+      "loss": 0.7116,
+      "step": 4931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4391075832989195,
+      "learning_rate": 1.0431439695856902e-05,
+      "loss": 0.7365,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7101197878193377,
+      "learning_rate": 1.0428148284701313e-05,
+      "loss": 0.7036,
+      "step": 4933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5331147672575676,
+      "learning_rate": 1.0424856827076926e-05,
+      "loss": 0.7613,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5142965480147195,
+      "learning_rate": 1.0421565323340971e-05,
+      "loss": 0.7013,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5435928907819016,
+      "learning_rate": 1.0418273773850696e-05,
+      "loss": 0.7334,
+      "step": 4936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.603917057655441,
+      "learning_rate": 1.0414982178963344e-05,
+      "loss": 0.6197,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6710661824730904,
+      "learning_rate": 1.0411690539036165e-05,
+      "loss": 0.7766,
+      "step": 4938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.796042940499753,
+      "learning_rate": 1.0408398854426413e-05,
+      "loss": 0.779,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5307692075330168,
+      "learning_rate": 1.0405107125491359e-05,
+      "loss": 0.7377,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7343403838596432,
+      "learning_rate": 1.0401815352588256e-05,
+      "loss": 0.6619,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6659604279911036,
+      "learning_rate": 1.0398523536074383e-05,
+      "loss": 0.7201,
+      "step": 4942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4924325225794244,
+      "learning_rate": 1.0395231676307012e-05,
+      "loss": 0.7449,
+      "step": 4943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6215899504993245,
+      "learning_rate": 1.039193977364342e-05,
+      "loss": 0.6166,
+      "step": 4944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5413755103496067,
+      "learning_rate": 1.0388647828440899e-05,
+      "loss": 0.6889,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8571969034985918,
+      "learning_rate": 1.0385355841056732e-05,
+      "loss": 0.7965,
+      "step": 4946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.505675486970467,
+      "learning_rate": 1.0382063811848216e-05,
+      "loss": 0.7368,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6573830089678232,
+      "learning_rate": 1.0378771741172647e-05,
+      "loss": 0.728,
+      "step": 4948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5151009935557034,
+      "learning_rate": 1.0375479629387328e-05,
+      "loss": 0.6786,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6888252027377988,
+      "learning_rate": 1.037218747684957e-05,
+      "loss": 0.785,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6060069581627165,
+      "learning_rate": 1.0368895283916678e-05,
+      "loss": 0.7782,
+      "step": 4951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5784964492645637,
+      "learning_rate": 1.0365603050945977e-05,
+      "loss": 0.6826,
+      "step": 4952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5425911773394898,
+      "learning_rate": 1.0362310778294775e-05,
+      "loss": 0.7056,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.599694320939845,
+      "learning_rate": 1.0359018466320407e-05,
+      "loss": 0.7094,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6737011916314462,
+      "learning_rate": 1.0355726115380203e-05,
+      "loss": 0.7664,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.530616756386857,
+      "learning_rate": 1.0352433725831487e-05,
+      "loss": 0.7243,
+      "step": 4956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5627773628239507,
+      "learning_rate": 1.0349141298031601e-05,
+      "loss": 0.6623,
+      "step": 4957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5693580207455513,
+      "learning_rate": 1.0345848832337891e-05,
+      "loss": 0.7476,
+      "step": 4958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4476294616961733,
+      "learning_rate": 1.0342556329107699e-05,
+      "loss": 0.7097,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4493711917271666,
+      "learning_rate": 1.0339263788698373e-05,
+      "loss": 0.607,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7351171237436391,
+      "learning_rate": 1.0335971211467265e-05,
+      "loss": 0.8264,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4803412656979784,
+      "learning_rate": 1.0332678597771739e-05,
+      "loss": 0.7487,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4887948440376384,
+      "learning_rate": 1.0329385947969154e-05,
+      "loss": 0.6569,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8106392631604589,
+      "learning_rate": 1.0326093262416874e-05,
+      "loss": 0.8789,
+      "step": 4964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6668317622828839,
+      "learning_rate": 1.0322800541472273e-05,
+      "loss": 0.8512,
+      "step": 4965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4499170830275672,
+      "learning_rate": 1.0319507785492718e-05,
+      "loss": 0.7089,
+      "step": 4966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4472378715645056,
+      "learning_rate": 1.031621499483559e-05,
+      "loss": 0.6131,
+      "step": 4967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6582927017623401,
+      "learning_rate": 1.031292216985827e-05,
+      "loss": 0.8072,
+      "step": 4968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4679495456025033,
+      "learning_rate": 1.030962931091814e-05,
+      "loss": 0.7026,
+      "step": 4969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6686120921027308,
+      "learning_rate": 1.0306336418372595e-05,
+      "loss": 0.6743,
+      "step": 4970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6297080468438931,
+      "learning_rate": 1.0303043492579017e-05,
+      "loss": 0.7175,
+      "step": 4971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4740594819956203,
+      "learning_rate": 1.029975053389481e-05,
+      "loss": 0.6553,
+      "step": 4972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6604044930548936,
+      "learning_rate": 1.0296457542677371e-05,
+      "loss": 0.7285,
+      "step": 4973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5331123784276655,
+      "learning_rate": 1.02931645192841e-05,
+      "loss": 0.6788,
+      "step": 4974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5496946285391504,
+      "learning_rate": 1.0289871464072405e-05,
+      "loss": 0.7427,
+      "step": 4975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5626346544494052,
+      "learning_rate": 1.0286578377399696e-05,
+      "loss": 0.5882,
+      "step": 4976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7064394550189277,
+      "learning_rate": 1.0283285259623385e-05,
+      "loss": 0.7806,
+      "step": 4977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.74754784601924,
+      "learning_rate": 1.0279992111100891e-05,
+      "loss": 0.8419,
+      "step": 4978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6513705967731058,
+      "learning_rate": 1.027669893218963e-05,
+      "loss": 0.7918,
+      "step": 4979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5863692267049108,
+      "learning_rate": 1.0273405723247028e-05,
+      "loss": 0.7655,
+      "step": 4980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5455744916739862,
+      "learning_rate": 1.027011248463051e-05,
+      "loss": 0.662,
+      "step": 4981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6189715089164287,
+      "learning_rate": 1.0266819216697506e-05,
+      "loss": 0.6878,
+      "step": 4982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5728852805180555,
+      "learning_rate": 1.0263525919805448e-05,
+      "loss": 0.7392,
+      "step": 4983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6485182074086644,
+      "learning_rate": 1.0260232594311775e-05,
+      "loss": 0.6494,
+      "step": 4984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6860934646556733,
+      "learning_rate": 1.025693924057392e-05,
+      "loss": 0.8304,
+      "step": 4985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5444623763479748,
+      "learning_rate": 1.025364585894933e-05,
+      "loss": 0.721,
+      "step": 4986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6114327837659634,
+      "learning_rate": 1.0250352449795449e-05,
+      "loss": 0.6927,
+      "step": 4987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6844313667842148,
+      "learning_rate": 1.0247059013469725e-05,
+      "loss": 0.8153,
+      "step": 4988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3912046210481805,
+      "learning_rate": 1.0243765550329606e-05,
+      "loss": 0.7573,
+      "step": 4989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7719163050138464,
+      "learning_rate": 1.0240472060732552e-05,
+      "loss": 0.7606,
+      "step": 4990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5907944102676055,
+      "learning_rate": 1.0237178545036015e-05,
+      "loss": 0.6463,
+      "step": 4991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4653612312225397,
+      "learning_rate": 1.0233885003597457e-05,
+      "loss": 0.6643,
+      "step": 4992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6555274215570521,
+      "learning_rate": 1.0230591436774337e-05,
+      "loss": 0.7665,
+      "step": 4993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.480369019040231,
+      "learning_rate": 1.0227297844924123e-05,
+      "loss": 0.7743,
+      "step": 4994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4834908923443815,
+      "learning_rate": 1.0224004228404286e-05,
+      "loss": 0.6652,
+      "step": 4995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7027958245467154,
+      "learning_rate": 1.0220710587572289e-05,
+      "loss": 0.6794,
+      "step": 4996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5892728457192657,
+      "learning_rate": 1.0217416922785613e-05,
+      "loss": 0.8259,
+      "step": 4997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.417842304139637,
+      "learning_rate": 1.0214123234401725e-05,
+      "loss": 0.6653,
+      "step": 4998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4897514214464702,
+      "learning_rate": 1.0210829522778111e-05,
+      "loss": 0.6722,
+      "step": 4999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4753853811493134,
+      "learning_rate": 1.020753578827225e-05,
+      "loss": 0.6329,
+      "step": 5000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.698061227933607,
+      "learning_rate": 1.0204242031241624e-05,
+      "loss": 0.6545,
+      "step": 5001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5685023415147041,
+      "learning_rate": 1.020094825204372e-05,
+      "loss": 0.7103,
+      "step": 5002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.757520070304479,
+      "learning_rate": 1.0197654451036025e-05,
+      "loss": 0.8271,
+      "step": 5003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.496912420552621,
+      "learning_rate": 1.0194360628576032e-05,
+      "loss": 0.646,
+      "step": 5004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6368808812363533,
+      "learning_rate": 1.0191066785021232e-05,
+      "loss": 0.6652,
+      "step": 5005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7515757375629755,
+      "learning_rate": 1.0187772920729118e-05,
+      "loss": 0.7075,
+      "step": 5006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.532702151487452,
+      "learning_rate": 1.0184479036057191e-05,
+      "loss": 0.7008,
+      "step": 5007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8067628392350867,
+      "learning_rate": 1.0181185131362949e-05,
+      "loss": 0.8778,
+      "step": 5008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5930483783286502,
+      "learning_rate": 1.0177891207003897e-05,
+      "loss": 0.6907,
+      "step": 5009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5893821059072393,
+      "learning_rate": 1.0174597263337533e-05,
+      "loss": 0.6614,
+      "step": 5010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6078386505119349,
+      "learning_rate": 1.0171303300721369e-05,
+      "loss": 0.8173,
+      "step": 5011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4627833807251032,
+      "learning_rate": 1.0168009319512908e-05,
+      "loss": 0.735,
+      "step": 5012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5924423835822419,
+      "learning_rate": 1.0164715320069667e-05,
+      "loss": 0.6657,
+      "step": 5013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3847413718790185,
+      "learning_rate": 1.016142130274915e-05,
+      "loss": 0.7492,
+      "step": 5014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4717912087381975,
+      "learning_rate": 1.0158127267908877e-05,
+      "loss": 0.6354,
+      "step": 5015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5804650771140725,
+      "learning_rate": 1.0154833215906359e-05,
+      "loss": 0.6814,
+      "step": 5016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.802943339801398,
+      "learning_rate": 1.015153914709912e-05,
+      "loss": 0.9026,
+      "step": 5017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5091332096941117,
+      "learning_rate": 1.0148245061844678e-05,
+      "loss": 0.8205,
+      "step": 5018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6077826529932098,
+      "learning_rate": 1.0144950960500548e-05,
+      "loss": 0.6848,
+      "step": 5019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.498282901072857,
+      "learning_rate": 1.0141656843424261e-05,
+      "loss": 0.7553,
+      "step": 5020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5345091773948873,
+      "learning_rate": 1.0138362710973337e-05,
+      "loss": 0.7252,
+      "step": 5021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5877742089090119,
+      "learning_rate": 1.0135068563505305e-05,
+      "loss": 0.6948,
+      "step": 5022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9732284318068651,
+      "learning_rate": 1.0131774401377694e-05,
+      "loss": 0.6836,
+      "step": 5023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4967048574718913,
+      "learning_rate": 1.0128480224948032e-05,
+      "loss": 0.653,
+      "step": 5024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.472468435443647,
+      "learning_rate": 1.0125186034573848e-05,
+      "loss": 0.7266,
+      "step": 5025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8378646799585223,
+      "learning_rate": 1.0121891830612682e-05,
+      "loss": 0.7543,
+      "step": 5026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7292167157582072,
+      "learning_rate": 1.0118597613422064e-05,
+      "loss": 0.777,
+      "step": 5027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.747577986929217,
+      "learning_rate": 1.0115303383359527e-05,
+      "loss": 0.6887,
+      "step": 5028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5050282607663057,
+      "learning_rate": 1.011200914078261e-05,
+      "loss": 0.7344,
+      "step": 5029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5261534192520534,
+      "learning_rate": 1.0108714886048854e-05,
+      "loss": 0.7532,
+      "step": 5030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4756222381550796,
+      "learning_rate": 1.0105420619515798e-05,
+      "loss": 0.7665,
+      "step": 5031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4293226846812288,
+      "learning_rate": 1.0102126341540981e-05,
+      "loss": 0.7232,
+      "step": 5032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8230140564560462,
+      "learning_rate": 1.0098832052481946e-05,
+      "loss": 0.7146,
+      "step": 5033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6393459817543277,
+      "learning_rate": 1.0095537752696238e-05,
+      "loss": 0.7186,
+      "step": 5034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.540375836450423,
+      "learning_rate": 1.0092243442541404e-05,
+      "loss": 0.6179,
+      "step": 5035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6002854156707615,
+      "learning_rate": 1.0088949122374982e-05,
+      "loss": 0.7583,
+      "step": 5036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.791900632212417,
+      "learning_rate": 1.0085654792554527e-05,
+      "loss": 0.6443,
+      "step": 5037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8762373825564007,
+      "learning_rate": 1.0082360453437583e-05,
+      "loss": 0.7761,
+      "step": 5038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.540714618809593,
+      "learning_rate": 1.00790661053817e-05,
+      "loss": 0.6828,
+      "step": 5039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.584376129187976,
+      "learning_rate": 1.0075771748744425e-05,
+      "loss": 0.739,
+      "step": 5040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8348523122891849,
+      "learning_rate": 1.0072477383883315e-05,
+      "loss": 0.6564,
+      "step": 5041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7931328635932917,
+      "learning_rate": 1.0069183011155919e-05,
+      "loss": 0.7407,
+      "step": 5042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6265720031446385,
+      "learning_rate": 1.0065888630919784e-05,
+      "loss": 0.8066,
+      "step": 5043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6535312579766697,
+      "learning_rate": 1.0062594243532473e-05,
+      "loss": 0.7392,
+      "step": 5044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6066458750468946,
+      "learning_rate": 1.0059299849351537e-05,
+      "loss": 0.772,
+      "step": 5045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.490332422068347,
+      "learning_rate": 1.0056005448734527e-05,
+      "loss": 0.7085,
+      "step": 5046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5434677098664784,
+      "learning_rate": 1.0052711042039e-05,
+      "loss": 0.6175,
+      "step": 5047
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7218251257628168,
+      "learning_rate": 1.0049416629622515e-05,
+      "loss": 0.7077,
+      "step": 5048
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5429660052228062,
+      "learning_rate": 1.0046122211842629e-05,
+      "loss": 0.6454,
+      "step": 5049
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7276557227894023,
+      "learning_rate": 1.0042827789056897e-05,
+      "loss": 0.8497,
+      "step": 5050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5107847951141293,
+      "learning_rate": 1.003953336162288e-05,
+      "loss": 0.675,
+      "step": 5051
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5534066484669686,
+      "learning_rate": 1.0036238929898133e-05,
+      "loss": 0.6476,
+      "step": 5052
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7502090023426649,
+      "learning_rate": 1.0032944494240215e-05,
+      "loss": 0.7414,
+      "step": 5053
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.61180441500941,
+      "learning_rate": 1.002965005500669e-05,
+      "loss": 0.7667,
+      "step": 5054
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5907909366442183,
+      "learning_rate": 1.0026355612555116e-05,
+      "loss": 0.8143,
+      "step": 5055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6927249074587167,
+      "learning_rate": 1.0023061167243051e-05,
+      "loss": 0.8298,
+      "step": 5056
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.558245964430676,
+      "learning_rate": 1.0019766719428056e-05,
+      "loss": 0.6458,
+      "step": 5057
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5989346457818998,
+      "learning_rate": 1.0016472269467694e-05,
+      "loss": 0.7566,
+      "step": 5058
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5501716886854584,
+      "learning_rate": 1.0013177817719525e-05,
+      "loss": 0.6956,
+      "step": 5059
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5298661053568097,
+      "learning_rate": 1.000988336454111e-05,
+      "loss": 0.7485,
+      "step": 5060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.534800077294969,
+      "learning_rate": 1.0006588910290009e-05,
+      "loss": 0.72,
+      "step": 5061
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4039336604523878,
+      "learning_rate": 1.0003294455323787e-05,
+      "loss": 0.6622,
+      "step": 5062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4625189066192077,
+      "learning_rate": 1e-05,
+      "loss": 0.7393,
+      "step": 5063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6009023241100437,
+      "learning_rate": 9.996705544676214e-06,
+      "loss": 0.7206,
+      "step": 5064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6802185162627696,
+      "learning_rate": 9.993411089709995e-06,
+      "loss": 0.703,
+      "step": 5065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4998292988494204,
+      "learning_rate": 9.990116635458893e-06,
+      "loss": 0.6667,
+      "step": 5066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5951409889363042,
+      "learning_rate": 9.98682218228048e-06,
+      "loss": 0.7688,
+      "step": 5067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4420550450229865,
+      "learning_rate": 9.98352773053231e-06,
+      "loss": 0.833,
+      "step": 5068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3540576285928896,
+      "learning_rate": 9.980233280571946e-06,
+      "loss": 0.6754,
+      "step": 5069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.58857712960787,
+      "learning_rate": 9.976938832756952e-06,
+      "loss": 0.7208,
+      "step": 5070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6888765345374839,
+      "learning_rate": 9.973644387444887e-06,
+      "loss": 0.8324,
+      "step": 5071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6628467015311343,
+      "learning_rate": 9.970349944993315e-06,
+      "loss": 0.7039,
+      "step": 5072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4986284311832823,
+      "learning_rate": 9.967055505759787e-06,
+      "loss": 0.6795,
+      "step": 5073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8380637247061966,
+      "learning_rate": 9.963761070101869e-06,
+      "loss": 0.8186,
+      "step": 5074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6236843675335488,
+      "learning_rate": 9.960466638377125e-06,
+      "loss": 0.7437,
+      "step": 5075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7043518992354776,
+      "learning_rate": 9.957172210943105e-06,
+      "loss": 0.6886,
+      "step": 5076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5516039920891835,
+      "learning_rate": 9.953877788157373e-06,
+      "loss": 0.7047,
+      "step": 5077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4788193377628533,
+      "learning_rate": 9.950583370377488e-06,
+      "loss": 0.7271,
+      "step": 5078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.560881796187408,
+      "learning_rate": 9.947288957961001e-06,
+      "loss": 0.6769,
+      "step": 5079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5307649077697028,
+      "learning_rate": 9.943994551265478e-06,
+      "loss": 0.6648,
+      "step": 5080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.565759677818178,
+      "learning_rate": 9.940700150648467e-06,
+      "loss": 0.7498,
+      "step": 5081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5830488858116634,
+      "learning_rate": 9.937405756467527e-06,
+      "loss": 0.7328,
+      "step": 5082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4985569654884294,
+      "learning_rate": 9.934111369080218e-06,
+      "loss": 0.6863,
+      "step": 5083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6931785992505062,
+      "learning_rate": 9.930816988844084e-06,
+      "loss": 0.6846,
+      "step": 5084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5202715907906006,
+      "learning_rate": 9.927522616116687e-06,
+      "loss": 0.6753,
+      "step": 5085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6362017411817606,
+      "learning_rate": 9.924228251255577e-06,
+      "loss": 0.8604,
+      "step": 5086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5759133874543885,
+      "learning_rate": 9.920933894618303e-06,
+      "loss": 0.696,
+      "step": 5087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7292168383139979,
+      "learning_rate": 9.91763954656242e-06,
+      "loss": 0.8085,
+      "step": 5088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5071322373168559,
+      "learning_rate": 9.914345207445476e-06,
+      "loss": 0.7611,
+      "step": 5089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6228730452289775,
+      "learning_rate": 9.911050877625022e-06,
+      "loss": 0.6774,
+      "step": 5090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.608877474252974,
+      "learning_rate": 9.9077565574586e-06,
+      "loss": 0.7135,
+      "step": 5091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5956742547959841,
+      "learning_rate": 9.904462247303764e-06,
+      "loss": 0.8134,
+      "step": 5092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4093253731337068,
+      "learning_rate": 9.901167947518057e-06,
+      "loss": 0.6145,
+      "step": 5093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6676622348956422,
+      "learning_rate": 9.89787365845902e-06,
+      "loss": 0.7878,
+      "step": 5094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.485830024047951,
+      "learning_rate": 9.894579380484206e-06,
+      "loss": 0.6392,
+      "step": 5095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8186613390608457,
+      "learning_rate": 9.89128511395115e-06,
+      "loss": 0.6955,
+      "step": 5096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5786168135399719,
+      "learning_rate": 9.887990859217391e-06,
+      "loss": 0.7995,
+      "step": 5097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7458285958219886,
+      "learning_rate": 9.88469661664048e-06,
+      "loss": 0.6816,
+      "step": 5098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.511674397329433,
+      "learning_rate": 9.881402386577942e-06,
+      "loss": 0.6325,
+      "step": 5099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5292384328734583,
+      "learning_rate": 9.87810816938732e-06,
+      "loss": 0.6494,
+      "step": 5100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4321640908065174,
+      "learning_rate": 9.874813965426154e-06,
+      "loss": 0.6423,
+      "step": 5101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.504524200960757,
+      "learning_rate": 9.871519775051971e-06,
+      "loss": 0.6557,
+      "step": 5102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4892683522137387,
+      "learning_rate": 9.86822559862231e-06,
+      "loss": 0.7091,
+      "step": 5103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6672876374415153,
+      "learning_rate": 9.864931436494696e-06,
+      "loss": 0.7504,
+      "step": 5104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5283640218290506,
+      "learning_rate": 9.861637289026664e-06,
+      "loss": 0.7191,
+      "step": 5105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.476362253308102,
+      "learning_rate": 9.858343156575744e-06,
+      "loss": 0.6484,
+      "step": 5106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6467227908333926,
+      "learning_rate": 9.855049039499454e-06,
+      "loss": 0.7415,
+      "step": 5107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5866704705759278,
+      "learning_rate": 9.851754938155329e-06,
+      "loss": 0.722,
+      "step": 5108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.51916605439026,
+      "learning_rate": 9.848460852900883e-06,
+      "loss": 0.6658,
+      "step": 5109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4627539145682227,
+      "learning_rate": 9.845166784093641e-06,
+      "loss": 0.675,
+      "step": 5110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.681030853676064,
+      "learning_rate": 9.841872732091128e-06,
+      "loss": 0.7634,
+      "step": 5111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7011084668318175,
+      "learning_rate": 9.838578697250852e-06,
+      "loss": 0.7477,
+      "step": 5112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6294344584210423,
+      "learning_rate": 9.835284679930335e-06,
+      "loss": 0.7684,
+      "step": 5113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6873612611256998,
+      "learning_rate": 9.831990680487094e-06,
+      "loss": 0.7013,
+      "step": 5114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5622192387852167,
+      "learning_rate": 9.828696699278633e-06,
+      "loss": 0.749,
+      "step": 5115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4396246048138015,
+      "learning_rate": 9.82540273666247e-06,
+      "loss": 0.6634,
+      "step": 5116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.59187032207698,
+      "learning_rate": 9.822108792996107e-06,
+      "loss": 0.7317,
+      "step": 5117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.573390113679101,
+      "learning_rate": 9.818814868637051e-06,
+      "loss": 0.7894,
+      "step": 5118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6061202138652004,
+      "learning_rate": 9.815520963942812e-06,
+      "loss": 0.6493,
+      "step": 5119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5454123702799032,
+      "learning_rate": 9.812227079270884e-06,
+      "loss": 0.7948,
+      "step": 5120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.597879788112768,
+      "learning_rate": 9.808933214978773e-06,
+      "loss": 0.7693,
+      "step": 5121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.690313982093511,
+      "learning_rate": 9.805639371423971e-06,
+      "loss": 0.6432,
+      "step": 5122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.677017710340723,
+      "learning_rate": 9.802345548963977e-06,
+      "loss": 0.7294,
+      "step": 5123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6484158834398428,
+      "learning_rate": 9.799051747956284e-06,
+      "loss": 0.718,
+      "step": 5124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7048099351617905,
+      "learning_rate": 9.795757968758378e-06,
+      "loss": 0.7263,
+      "step": 5125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6467261635402397,
+      "learning_rate": 9.792464211727755e-06,
+      "loss": 0.6859,
+      "step": 5126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3529923691945207,
+      "learning_rate": 9.789170477221892e-06,
+      "loss": 0.6789,
+      "step": 5127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.608164579433592,
+      "learning_rate": 9.785876765598276e-06,
+      "loss": 0.754,
+      "step": 5128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6396310055821297,
+      "learning_rate": 9.782583077214394e-06,
+      "loss": 0.7712,
+      "step": 5129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6564971675287619,
+      "learning_rate": 9.779289412427714e-06,
+      "loss": 0.7607,
+      "step": 5130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6173430588878457,
+      "learning_rate": 9.775995771595716e-06,
+      "loss": 0.7981,
+      "step": 5131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6017258418261529,
+      "learning_rate": 9.772702155075879e-06,
+      "loss": 0.7253,
+      "step": 5132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7709026391696094,
+      "learning_rate": 9.769408563225665e-06,
+      "loss": 0.7332,
+      "step": 5133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.519283381436825,
+      "learning_rate": 9.76611499640255e-06,
+      "loss": 0.6772,
+      "step": 5134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.609935659876278,
+      "learning_rate": 9.76282145496399e-06,
+      "loss": 0.7068,
+      "step": 5135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.531593684480193,
+      "learning_rate": 9.75952793926745e-06,
+      "loss": 0.646,
+      "step": 5136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5790104519401773,
+      "learning_rate": 9.756234449670396e-06,
+      "loss": 0.6888,
+      "step": 5137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.446800963186642,
+      "learning_rate": 9.752940986530279e-06,
+      "loss": 0.6697,
+      "step": 5138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6523518022610042,
+      "learning_rate": 9.749647550204555e-06,
+      "loss": 0.7867,
+      "step": 5139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4971052938564537,
+      "learning_rate": 9.746354141050674e-06,
+      "loss": 0.7876,
+      "step": 5140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.570377547663415,
+      "learning_rate": 9.743060759426082e-06,
+      "loss": 0.7654,
+      "step": 5141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6064579733110664,
+      "learning_rate": 9.739767405688228e-06,
+      "loss": 0.737,
+      "step": 5142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5455426190649941,
+      "learning_rate": 9.736474080194555e-06,
+      "loss": 0.6627,
+      "step": 5143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7875043911265842,
+      "learning_rate": 9.733180783302499e-06,
+      "loss": 0.8372,
+      "step": 5144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4778229053041372,
+      "learning_rate": 9.729887515369491e-06,
+      "loss": 0.6284,
+      "step": 5145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6191155801176957,
+      "learning_rate": 9.726594276752975e-06,
+      "loss": 0.663,
+      "step": 5146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8209617158366576,
+      "learning_rate": 9.723301067810373e-06,
+      "loss": 0.8653,
+      "step": 5147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6978790968854747,
+      "learning_rate": 9.720007888899112e-06,
+      "loss": 0.7375,
+      "step": 5148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5651878518597222,
+      "learning_rate": 9.716714740376616e-06,
+      "loss": 0.7489,
+      "step": 5149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2843081750148266,
+      "learning_rate": 9.713421622600307e-06,
+      "loss": 0.5879,
+      "step": 5150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.553289187108251,
+      "learning_rate": 9.710128535927597e-06,
+      "loss": 0.7157,
+      "step": 5151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6699906188531186,
+      "learning_rate": 9.706835480715905e-06,
+      "loss": 0.8308,
+      "step": 5152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.512597982088698,
+      "learning_rate": 9.703542457322632e-06,
+      "loss": 0.7099,
+      "step": 5153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.416985900975218,
+      "learning_rate": 9.700249466105193e-06,
+      "loss": 0.6356,
+      "step": 5154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6035167743253373,
+      "learning_rate": 9.696956507420984e-06,
+      "loss": 0.6669,
+      "step": 5155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.534607483177831,
+      "learning_rate": 9.693663581627408e-06,
+      "loss": 0.6586,
+      "step": 5156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5302169037608306,
+      "learning_rate": 9.690370689081863e-06,
+      "loss": 0.674,
+      "step": 5157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6115188616385638,
+      "learning_rate": 9.687077830141734e-06,
+      "loss": 0.6778,
+      "step": 5158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.618571618506667,
+      "learning_rate": 9.683785005164412e-06,
+      "loss": 0.7339,
+      "step": 5159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4804908116816398,
+      "learning_rate": 9.680492214507287e-06,
+      "loss": 0.6456,
+      "step": 5160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.494140203548696,
+      "learning_rate": 9.67719945852773e-06,
+      "loss": 0.717,
+      "step": 5161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6607958643733662,
+      "learning_rate": 9.67390673758313e-06,
+      "loss": 0.7738,
+      "step": 5162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4974459021451776,
+      "learning_rate": 9.670614052030849e-06,
+      "loss": 0.8225,
+      "step": 5163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7055403350977216,
+      "learning_rate": 9.667321402228261e-06,
+      "loss": 0.7338,
+      "step": 5164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5909879761244057,
+      "learning_rate": 9.664028788532738e-06,
+      "loss": 0.7313,
+      "step": 5165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5359287895843046,
+      "learning_rate": 9.66073621130163e-06,
+      "loss": 0.8177,
+      "step": 5166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.503526554781987,
+      "learning_rate": 9.657443670892303e-06,
+      "loss": 0.6308,
+      "step": 5167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6412671598656263,
+      "learning_rate": 9.65415116766211e-06,
+      "loss": 0.6373,
+      "step": 5168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5027991930993563,
+      "learning_rate": 9.650858701968399e-06,
+      "loss": 0.666,
+      "step": 5169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5524957943245377,
+      "learning_rate": 9.647566274168516e-06,
+      "loss": 0.681,
+      "step": 5170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4380181512787078,
+      "learning_rate": 9.644273884619802e-06,
+      "loss": 0.7568,
+      "step": 5171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6762040745449318,
+      "learning_rate": 9.640981533679593e-06,
+      "loss": 0.6954,
+      "step": 5172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.592284969495283,
+      "learning_rate": 9.637689221705227e-06,
+      "loss": 0.6853,
+      "step": 5173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4948404765759724,
+      "learning_rate": 9.634396949054028e-06,
+      "loss": 0.6828,
+      "step": 5174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5866449703821197,
+      "learning_rate": 9.631104716083327e-06,
+      "loss": 0.6721,
+      "step": 5175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4904129492677478,
+      "learning_rate": 9.627812523150434e-06,
+      "loss": 0.6681,
+      "step": 5176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6687105576410526,
+      "learning_rate": 9.624520370612672e-06,
+      "loss": 0.7614,
+      "step": 5177
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.662810354791064,
+      "learning_rate": 9.621228258827358e-06,
+      "loss": 0.6917,
+      "step": 5178
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.597033737761668,
+      "learning_rate": 9.617936188151786e-06,
+      "loss": 0.6599,
+      "step": 5179
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.59732571462882,
+      "learning_rate": 9.614644158943273e-06,
+      "loss": 0.7304,
+      "step": 5180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4910868084236737,
+      "learning_rate": 9.611352171559105e-06,
+      "loss": 0.6728,
+      "step": 5181
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.608341256608166,
+      "learning_rate": 9.60806022635658e-06,
+      "loss": 0.7524,
+      "step": 5182
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.550887296530815,
+      "learning_rate": 9.604768323692993e-06,
+      "loss": 0.6347,
+      "step": 5183
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7808332484032412,
+      "learning_rate": 9.601476463925619e-06,
+      "loss": 0.8607,
+      "step": 5184
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.556241721735327,
+      "learning_rate": 9.598184647411744e-06,
+      "loss": 0.6297,
+      "step": 5185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.585860373624063,
+      "learning_rate": 9.594892874508646e-06,
+      "loss": 0.6625,
+      "step": 5186
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6386386267965907,
+      "learning_rate": 9.591601145573585e-06,
+      "loss": 0.7331,
+      "step": 5187
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4186691498174373,
+      "learning_rate": 9.58830946096384e-06,
+      "loss": 0.6774,
+      "step": 5188
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5007796466653354,
+      "learning_rate": 9.58501782103666e-06,
+      "loss": 0.789,
+      "step": 5189
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6658588544515636,
+      "learning_rate": 9.581726226149304e-06,
+      "loss": 0.6567,
+      "step": 5190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4548246999305345,
+      "learning_rate": 9.57843467665903e-06,
+      "loss": 0.7615,
+      "step": 5191
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.487561384764103,
+      "learning_rate": 9.575143172923076e-06,
+      "loss": 0.6926,
+      "step": 5192
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.583406569115701,
+      "learning_rate": 9.571851715298688e-06,
+      "loss": 0.6854,
+      "step": 5193
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7058952396836893,
+      "learning_rate": 9.568560304143101e-06,
+      "loss": 0.7685,
+      "step": 5194
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4557818774250706,
+      "learning_rate": 9.565268939813545e-06,
+      "loss": 0.6032,
+      "step": 5195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5812853823646884,
+      "learning_rate": 9.561977622667248e-06,
+      "loss": 0.6898,
+      "step": 5196
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5803499566349357,
+      "learning_rate": 9.55868635306143e-06,
+      "loss": 0.7041,
+      "step": 5197
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5179266574469583,
+      "learning_rate": 9.555395131353307e-06,
+      "loss": 0.6686,
+      "step": 5198
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.583336365752639,
+      "learning_rate": 9.55210395790009e-06,
+      "loss": 0.7714,
+      "step": 5199
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7268133639594228,
+      "learning_rate": 9.548812833058984e-06,
+      "loss": 0.7094,
+      "step": 5200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5994523133476826,
+      "learning_rate": 9.545521757187188e-06,
+      "loss": 0.6845,
+      "step": 5201
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5297856828233163,
+      "learning_rate": 9.542230730641898e-06,
+      "loss": 0.689,
+      "step": 5202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6519608888054835,
+      "learning_rate": 9.538939753780304e-06,
+      "loss": 0.7541,
+      "step": 5203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5538051540501847,
+      "learning_rate": 9.535648826959591e-06,
+      "loss": 0.7378,
+      "step": 5204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5560502994532663,
+      "learning_rate": 9.532357950536934e-06,
+      "loss": 0.7586,
+      "step": 5205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5562973234346362,
+      "learning_rate": 9.529067124869512e-06,
+      "loss": 0.7769,
+      "step": 5206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4718150268643992,
+      "learning_rate": 9.525776350314486e-06,
+      "loss": 0.6668,
+      "step": 5207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5758318544958667,
+      "learning_rate": 9.522485627229022e-06,
+      "loss": 0.7326,
+      "step": 5208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6178121744079512,
+      "learning_rate": 9.519194955970275e-06,
+      "loss": 0.7513,
+      "step": 5209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5169507449367376,
+      "learning_rate": 9.515904336895395e-06,
+      "loss": 0.7525,
+      "step": 5210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4632947803615415,
+      "learning_rate": 9.512613770361532e-06,
+      "loss": 0.6774,
+      "step": 5211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5564286603593986,
+      "learning_rate": 9.50932325672582e-06,
+      "loss": 0.7545,
+      "step": 5212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4108790343983597,
+      "learning_rate": 9.506032796345394e-06,
+      "loss": 0.6761,
+      "step": 5213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4547297896795017,
+      "learning_rate": 9.50274238957739e-06,
+      "loss": 0.7189,
+      "step": 5214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6395134721925821,
+      "learning_rate": 9.499452036778917e-06,
+      "loss": 0.6663,
+      "step": 5215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4346935198396273,
+      "learning_rate": 9.4961617383071e-06,
+      "loss": 0.5815,
+      "step": 5216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5006255850588064,
+      "learning_rate": 9.492871494519047e-06,
+      "loss": 0.7279,
+      "step": 5217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4022924331040245,
+      "learning_rate": 9.48958130577186e-06,
+      "loss": 0.6135,
+      "step": 5218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7138990478244631,
+      "learning_rate": 9.486291172422646e-06,
+      "loss": 0.8144,
+      "step": 5219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6216852067769545,
+      "learning_rate": 9.483001094828486e-06,
+      "loss": 0.7125,
+      "step": 5220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5811436084819726,
+      "learning_rate": 9.479711073346477e-06,
+      "loss": 0.605,
+      "step": 5221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5977073093399896,
+      "learning_rate": 9.476421108333692e-06,
+      "loss": 0.6632,
+      "step": 5222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6516495394418902,
+      "learning_rate": 9.473131200147205e-06,
+      "loss": 0.6504,
+      "step": 5223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4696334684470354,
+      "learning_rate": 9.469841349144092e-06,
+      "loss": 0.6536,
+      "step": 5224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.398855086381696,
+      "learning_rate": 9.466551555681406e-06,
+      "loss": 0.7457,
+      "step": 5225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7310787645704016,
+      "learning_rate": 9.463261820116207e-06,
+      "loss": 0.7888,
+      "step": 5226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4237314465626654,
+      "learning_rate": 9.459972142805546e-06,
+      "loss": 0.8367,
+      "step": 5227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7246104531629485,
+      "learning_rate": 9.456682524106462e-06,
+      "loss": 0.7903,
+      "step": 5228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5646826279558628,
+      "learning_rate": 9.453392964375997e-06,
+      "loss": 0.7519,
+      "step": 5229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5724704042349023,
+      "learning_rate": 9.450103463971173e-06,
+      "loss": 0.6894,
+      "step": 5230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.563277249429908,
+      "learning_rate": 9.446814023249017e-06,
+      "loss": 0.7203,
+      "step": 5231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6000393194900262,
+      "learning_rate": 9.443524642566555e-06,
+      "loss": 0.704,
+      "step": 5232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5404602459582803,
+      "learning_rate": 9.440235322280787e-06,
+      "loss": 0.6584,
+      "step": 5233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.731041243050856,
+      "learning_rate": 9.436946062748723e-06,
+      "loss": 0.6915,
+      "step": 5234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4755425913249085,
+      "learning_rate": 9.433656864327355e-06,
+      "loss": 0.7044,
+      "step": 5235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.630038704091692,
+      "learning_rate": 9.430367727373677e-06,
+      "loss": 0.6958,
+      "step": 5236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5067240174520171,
+      "learning_rate": 9.42707865224468e-06,
+      "loss": 0.6853,
+      "step": 5237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4613411878535234,
+      "learning_rate": 9.42378963929733e-06,
+      "loss": 0.7166,
+      "step": 5238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4578198906412034,
+      "learning_rate": 9.42050068888861e-06,
+      "loss": 0.6512,
+      "step": 5239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7000464098217933,
+      "learning_rate": 9.417211801375473e-06,
+      "loss": 0.8523,
+      "step": 5240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4453794950143881,
+      "learning_rate": 9.41392297711488e-06,
+      "loss": 0.7007,
+      "step": 5241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7221610086703045,
+      "learning_rate": 9.410634216463788e-06,
+      "loss": 0.7637,
+      "step": 5242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.523370062055753,
+      "learning_rate": 9.407345519779129e-06,
+      "loss": 0.6814,
+      "step": 5243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.608673973036898,
+      "learning_rate": 9.404056887417847e-06,
+      "loss": 0.65,
+      "step": 5244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6451242028947322,
+      "learning_rate": 9.400768319736874e-06,
+      "loss": 0.694,
+      "step": 5245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6038945321260987,
+      "learning_rate": 9.397479817093126e-06,
+      "loss": 0.6588,
+      "step": 5246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6994527008054192,
+      "learning_rate": 9.394191379843524e-06,
+      "loss": 0.8255,
+      "step": 5247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.681999020364834,
+      "learning_rate": 9.390903008344972e-06,
+      "loss": 0.6967,
+      "step": 5248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5577270594188073,
+      "learning_rate": 9.387614702954372e-06,
+      "loss": 0.6311,
+      "step": 5249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6413677367230959,
+      "learning_rate": 9.384326464028622e-06,
+      "loss": 0.7477,
+      "step": 5250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5642001057958752,
+      "learning_rate": 9.381038291924607e-06,
+      "loss": 0.733,
+      "step": 5251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.505431451576894,
+      "learning_rate": 9.377750186999203e-06,
+      "loss": 0.6696,
+      "step": 5252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5609368395988965,
+      "learning_rate": 9.374462149609286e-06,
+      "loss": 0.7326,
+      "step": 5253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4406046445318315,
+      "learning_rate": 9.371174180111722e-06,
+      "loss": 0.6684,
+      "step": 5254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6830633570580327,
+      "learning_rate": 9.367886278863366e-06,
+      "loss": 0.7631,
+      "step": 5255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4375200749846495,
+      "learning_rate": 9.364598446221068e-06,
+      "loss": 0.7355,
+      "step": 5256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5440091833277783,
+      "learning_rate": 9.361310682541675e-06,
+      "loss": 0.7658,
+      "step": 5257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.352238801113515,
+      "learning_rate": 9.358022988182017e-06,
+      "loss": 0.6368,
+      "step": 5258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5949685287350601,
+      "learning_rate": 9.354735363498925e-06,
+      "loss": 0.709,
+      "step": 5259
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4948494174337732,
+      "learning_rate": 9.351447808849219e-06,
+      "loss": 0.6661,
+      "step": 5260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.624657505673658,
+      "learning_rate": 9.348160324589709e-06,
+      "loss": 0.7397,
+      "step": 5261
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5507950151661274,
+      "learning_rate": 9.344872911077206e-06,
+      "loss": 0.7714,
+      "step": 5262
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5574701634407244,
+      "learning_rate": 9.341585568668501e-06,
+      "loss": 0.7168,
+      "step": 5263
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4886619135598378,
+      "learning_rate": 9.338298297720385e-06,
+      "loss": 0.7643,
+      "step": 5264
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.48937824246935,
+      "learning_rate": 9.335011098589644e-06,
+      "loss": 0.7108,
+      "step": 5265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5537164166555901,
+      "learning_rate": 9.331723971633046e-06,
+      "loss": 0.6894,
+      "step": 5266
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4409876623283242,
+      "learning_rate": 9.328436917207359e-06,
+      "loss": 0.6557,
+      "step": 5267
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3906556675122819,
+      "learning_rate": 9.325149935669347e-06,
+      "loss": 0.6577,
+      "step": 5268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4896382181245778,
+      "learning_rate": 9.321863027375753e-06,
+      "loss": 0.6772,
+      "step": 5269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5065184254475665,
+      "learning_rate": 9.318576192683326e-06,
+      "loss": 0.6624,
+      "step": 5270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6122106573181099,
+      "learning_rate": 9.315289431948794e-06,
+      "loss": 0.7075,
+      "step": 5271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6082453926127243,
+      "learning_rate": 9.312002745528885e-06,
+      "loss": 0.7192,
+      "step": 5272
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5519173002255817,
+      "learning_rate": 9.308716133780323e-06,
+      "loss": 0.7058,
+      "step": 5273
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5511865568301473,
+      "learning_rate": 9.30542959705981e-06,
+      "loss": 0.6688,
+      "step": 5274
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6629909624506398,
+      "learning_rate": 9.302143135724058e-06,
+      "loss": 0.722,
+      "step": 5275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.554510834086878,
+      "learning_rate": 9.29885675012975e-06,
+      "loss": 0.7647,
+      "step": 5276
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4276555624226124,
+      "learning_rate": 9.295570440633577e-06,
+      "loss": 0.6158,
+      "step": 5277
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.741157099895649,
+      "learning_rate": 9.29228420759222e-06,
+      "loss": 0.6786,
+      "step": 5278
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.619832980822339,
+      "learning_rate": 9.288998051362343e-06,
+      "loss": 0.7473,
+      "step": 5279
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3846762393519525,
+      "learning_rate": 9.285711972300607e-06,
+      "loss": 0.5589,
+      "step": 5280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.494543803188233,
+      "learning_rate": 9.28242597076367e-06,
+      "loss": 0.6714,
+      "step": 5281
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.54481372971172,
+      "learning_rate": 9.27914004710817e-06,
+      "loss": 0.716,
+      "step": 5282
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5701962361404953,
+      "learning_rate": 9.275854201690748e-06,
+      "loss": 0.7484,
+      "step": 5283
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.564642596041506,
+      "learning_rate": 9.272568434868024e-06,
+      "loss": 0.6589,
+      "step": 5284
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5127112390691917,
+      "learning_rate": 9.26928274699662e-06,
+      "loss": 0.692,
+      "step": 5285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6438483898007703,
+      "learning_rate": 9.265997138433152e-06,
+      "loss": 0.7135,
+      "step": 5286
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.418691263372242,
+      "learning_rate": 9.262711609534211e-06,
+      "loss": 0.6108,
+      "step": 5287
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4353502362260733,
+      "learning_rate": 9.2594261606564e-06,
+      "loss": 0.6181,
+      "step": 5288
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5773405361749253,
+      "learning_rate": 9.256140792156292e-06,
+      "loss": 0.6366,
+      "step": 5289
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.740700222032807,
+      "learning_rate": 9.25285550439047e-06,
+      "loss": 0.713,
+      "step": 5290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5815623965705798,
+      "learning_rate": 9.249570297715503e-06,
+      "loss": 0.6745,
+      "step": 5291
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7360353904593255,
+      "learning_rate": 9.24628517248794e-06,
+      "loss": 0.7386,
+      "step": 5292
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.600154153236615,
+      "learning_rate": 9.243000129064339e-06,
+      "loss": 0.7301,
+      "step": 5293
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6322114591649122,
+      "learning_rate": 9.239715167801232e-06,
+      "loss": 0.6683,
+      "step": 5294
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5597922373861357,
+      "learning_rate": 9.236430289055154e-06,
+      "loss": 0.7467,
+      "step": 5295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5373445772916388,
+      "learning_rate": 9.23314549318263e-06,
+      "loss": 0.7699,
+      "step": 5296
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.480216056117163,
+      "learning_rate": 9.229860780540168e-06,
+      "loss": 0.7339,
+      "step": 5297
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6150409063359652,
+      "learning_rate": 9.226576151484274e-06,
+      "loss": 0.7056,
+      "step": 5298
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5014317259723393,
+      "learning_rate": 9.223291606371448e-06,
+      "loss": 0.7125,
+      "step": 5299
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4923615606329292,
+      "learning_rate": 9.220007145558169e-06,
+      "loss": 0.6393,
+      "step": 5300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5687143048914496,
+      "learning_rate": 9.216722769400917e-06,
+      "loss": 0.6065,
+      "step": 5301
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5118993852348217,
+      "learning_rate": 9.213438478256159e-06,
+      "loss": 0.6034,
+      "step": 5302
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4038497103453222,
+      "learning_rate": 9.210154272480353e-06,
+      "loss": 0.7255,
+      "step": 5303
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7505356088247321,
+      "learning_rate": 9.20687015242995e-06,
+      "loss": 0.6883,
+      "step": 5304
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.552647846292909,
+      "learning_rate": 9.203586118461389e-06,
+      "loss": 0.6346,
+      "step": 5305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7016995825452594,
+      "learning_rate": 9.2003021709311e-06,
+      "loss": 0.6711,
+      "step": 5306
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7047703184824687,
+      "learning_rate": 9.197018310195503e-06,
+      "loss": 0.7431,
+      "step": 5307
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6009774451727634,
+      "learning_rate": 9.193734536611013e-06,
+      "loss": 0.753,
+      "step": 5308
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5727984936944126,
+      "learning_rate": 9.19045085053403e-06,
+      "loss": 0.6955,
+      "step": 5309
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5171645830514315,
+      "learning_rate": 9.18716725232095e-06,
+      "loss": 0.6256,
+      "step": 5310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5013686068549332,
+      "learning_rate": 9.18388374232815e-06,
+      "loss": 0.7419,
+      "step": 5311
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5452514446664825,
+      "learning_rate": 9.180600320912006e-06,
+      "loss": 0.8011,
+      "step": 5312
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6779607379171213,
+      "learning_rate": 9.177316988428885e-06,
+      "loss": 0.7006,
+      "step": 5313
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5460751326931226,
+      "learning_rate": 9.174033745235141e-06,
+      "loss": 0.726,
+      "step": 5314
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4406243469055928,
+      "learning_rate": 9.170750591687114e-06,
+      "loss": 0.7465,
+      "step": 5315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6337337025110672,
+      "learning_rate": 9.167467528141144e-06,
+      "loss": 0.7255,
+      "step": 5316
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5467283757292964,
+      "learning_rate": 9.164184554953557e-06,
+      "loss": 0.7321,
+      "step": 5317
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.682046371729968,
+      "learning_rate": 9.160901672480662e-06,
+      "loss": 0.6444,
+      "step": 5318
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.488118046274293,
+      "learning_rate": 9.157618881078772e-06,
+      "loss": 0.713,
+      "step": 5319
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7079121163396311,
+      "learning_rate": 9.154336181104175e-06,
+      "loss": 0.6949,
+      "step": 5320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5068370655989956,
+      "learning_rate": 9.151053572913163e-06,
+      "loss": 0.7095,
+      "step": 5321
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7707562315922247,
+      "learning_rate": 9.147771056862011e-06,
+      "loss": 0.7966,
+      "step": 5322
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6133590922465229,
+      "learning_rate": 9.144488633306981e-06,
+      "loss": 0.6659,
+      "step": 5323
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4545510638629,
+      "learning_rate": 9.141206302604336e-06,
+      "loss": 0.7483,
+      "step": 5324
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.655522474567456,
+      "learning_rate": 9.13792406511031e-06,
+      "loss": 0.76,
+      "step": 5325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3576656692724869,
+      "learning_rate": 9.134641921181147e-06,
+      "loss": 0.6448,
+      "step": 5326
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7787458227500046,
+      "learning_rate": 9.131359871173074e-06,
+      "loss": 0.7244,
+      "step": 5327
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6178083657835598,
+      "learning_rate": 9.1280779154423e-06,
+      "loss": 0.6684,
+      "step": 5328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7010524693762994,
+      "learning_rate": 9.124796054345036e-06,
+      "loss": 0.7417,
+      "step": 5329
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5360912007131207,
+      "learning_rate": 9.12151428823747e-06,
+      "loss": 0.7682,
+      "step": 5330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7481181530838306,
+      "learning_rate": 9.118232617475791e-06,
+      "loss": 0.7718,
+      "step": 5331
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5760176359497586,
+      "learning_rate": 9.114951042416175e-06,
+      "loss": 0.6897,
+      "step": 5332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4532079356845804,
+      "learning_rate": 9.11166956341478e-06,
+      "loss": 0.7062,
+      "step": 5333
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5295195035357474,
+      "learning_rate": 9.10838818082776e-06,
+      "loss": 0.6727,
+      "step": 5334
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.502411837379029,
+      "learning_rate": 9.105106895011263e-06,
+      "loss": 0.6246,
+      "step": 5335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6815569575078204,
+      "learning_rate": 9.101825706321417e-06,
+      "loss": 0.6468,
+      "step": 5336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6571536250194325,
+      "learning_rate": 9.098544615114347e-06,
+      "loss": 0.7936,
+      "step": 5337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5541916532325835,
+      "learning_rate": 9.095263621746159e-06,
+      "loss": 0.7494,
+      "step": 5338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5773470344567357,
+      "learning_rate": 9.091982726572956e-06,
+      "loss": 0.7626,
+      "step": 5339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7006352160254827,
+      "learning_rate": 9.088701929950831e-06,
+      "loss": 0.6884,
+      "step": 5340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5057538412071996,
+      "learning_rate": 9.085421232235857e-06,
+      "loss": 0.5913,
+      "step": 5341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.681665813385068,
+      "learning_rate": 9.08214063378411e-06,
+      "loss": 0.8246,
+      "step": 5342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.616750287402683,
+      "learning_rate": 9.07886013495164e-06,
+      "loss": 0.6381,
+      "step": 5343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5931263201795647,
+      "learning_rate": 9.075579736094497e-06,
+      "loss": 0.6813,
+      "step": 5344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5711422910918862,
+      "learning_rate": 9.072299437568722e-06,
+      "loss": 0.7396,
+      "step": 5345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4722021442670385,
+      "learning_rate": 9.069019239730329e-06,
+      "loss": 0.7187,
+      "step": 5346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.684898911516207,
+      "learning_rate": 9.065739142935345e-06,
+      "loss": 0.784,
+      "step": 5347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5649833582235908,
+      "learning_rate": 9.062459147539761e-06,
+      "loss": 0.7162,
+      "step": 5348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.70568458464764,
+      "learning_rate": 9.059179253899577e-06,
+      "loss": 0.6717,
+      "step": 5349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.516662532354538,
+      "learning_rate": 9.055899462370776e-06,
+      "loss": 0.6827,
+      "step": 5350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.527811050777723,
+      "learning_rate": 9.052619773309318e-06,
+      "loss": 0.715,
+      "step": 5351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.394315245520379,
+      "learning_rate": 9.04934018707117e-06,
+      "loss": 0.6611,
+      "step": 5352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.669341562664938,
+      "learning_rate": 9.046060704012281e-06,
+      "loss": 0.7238,
+      "step": 5353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4150418929513942,
+      "learning_rate": 9.042781324488582e-06,
+      "loss": 0.6455,
+      "step": 5354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7234502937842522,
+      "learning_rate": 9.039502048856002e-06,
+      "loss": 0.7208,
+      "step": 5355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4593406939103721,
+      "learning_rate": 9.036222877470455e-06,
+      "loss": 0.693,
+      "step": 5356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7215012614015754,
+      "learning_rate": 9.032943810687841e-06,
+      "loss": 0.792,
+      "step": 5357
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6688124560243105,
+      "learning_rate": 9.029664848864055e-06,
+      "loss": 0.6758,
+      "step": 5358
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8236526409515186,
+      "learning_rate": 9.026385992354974e-06,
+      "loss": 0.7684,
+      "step": 5359
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.633138913316259,
+      "learning_rate": 9.023107241516469e-06,
+      "loss": 0.7151,
+      "step": 5360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5803702477774828,
+      "learning_rate": 9.019828596704394e-06,
+      "loss": 0.8013,
+      "step": 5361
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.669277721960774,
+      "learning_rate": 9.016550058274599e-06,
+      "loss": 0.7613,
+      "step": 5362
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5690020797998012,
+      "learning_rate": 9.013271626582915e-06,
+      "loss": 0.7337,
+      "step": 5363
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.41272139408782,
+      "learning_rate": 9.009993301985167e-06,
+      "loss": 0.6609,
+      "step": 5364
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.468996080511503,
+      "learning_rate": 9.006715084837162e-06,
+      "loss": 0.5811,
+      "step": 5365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5838321250978402,
+      "learning_rate": 9.003436975494699e-06,
+      "loss": 0.6922,
+      "step": 5366
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4850631249125565,
+      "learning_rate": 9.00015897431357e-06,
+      "loss": 0.6375,
+      "step": 5367
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.529836227289035,
+      "learning_rate": 8.996881081649551e-06,
+      "loss": 0.629,
+      "step": 5368
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5638270705319934,
+      "learning_rate": 8.9936032978584e-06,
+      "loss": 0.636,
+      "step": 5369
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4665776471013017,
+      "learning_rate": 8.990325623295877e-06,
+      "loss": 0.7205,
+      "step": 5370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.651673457610251,
+      "learning_rate": 8.987048058317714e-06,
+      "loss": 0.7849,
+      "step": 5371
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7829197596581683,
+      "learning_rate": 8.983770603279642e-06,
+      "loss": 0.751,
+      "step": 5372
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8456608183651604,
+      "learning_rate": 8.980493258537384e-06,
+      "loss": 0.7839,
+      "step": 5373
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.703926959579352,
+      "learning_rate": 8.977216024446636e-06,
+      "loss": 0.8251,
+      "step": 5374
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.530470230657618,
+      "learning_rate": 8.973938901363094e-06,
+      "loss": 0.6577,
+      "step": 5375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5809578736817425,
+      "learning_rate": 8.970661889642442e-06,
+      "loss": 0.7201,
+      "step": 5376
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4271102652547045,
+      "learning_rate": 8.96738498964034e-06,
+      "loss": 0.6792,
+      "step": 5377
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5426396797757143,
+      "learning_rate": 8.964108201712455e-06,
+      "loss": 0.6958,
+      "step": 5378
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8103751233562824,
+      "learning_rate": 8.96083152621442e-06,
+      "loss": 0.8033,
+      "step": 5379
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5761100242991692,
+      "learning_rate": 8.957554963501873e-06,
+      "loss": 0.7647,
+      "step": 5380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5905942042313983,
+      "learning_rate": 8.954278513930434e-06,
+      "loss": 0.7884,
+      "step": 5381
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5539806558481204,
+      "learning_rate": 8.951002177855708e-06,
+      "loss": 0.7533,
+      "step": 5382
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6703501138543935,
+      "learning_rate": 8.947725955633294e-06,
+      "loss": 0.749,
+      "step": 5383
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5119616474184723,
+      "learning_rate": 8.944449847618768e-06,
+      "loss": 0.6985,
+      "step": 5384
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6205786635233608,
+      "learning_rate": 8.941173854167703e-06,
+      "loss": 0.7124,
+      "step": 5385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5975066009074679,
+      "learning_rate": 8.937897975635664e-06,
+      "loss": 0.6692,
+      "step": 5386
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.636521080970513,
+      "learning_rate": 8.934622212378185e-06,
+      "loss": 0.6629,
+      "step": 5387
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5190343789945384,
+      "learning_rate": 8.931346564750808e-06,
+      "loss": 0.651,
+      "step": 5388
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5957467566016765,
+      "learning_rate": 8.928071033109047e-06,
+      "loss": 0.7199,
+      "step": 5389
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4624387519553008,
+      "learning_rate": 8.924795617808411e-06,
+      "loss": 0.6886,
+      "step": 5390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4511607366742325,
+      "learning_rate": 8.921520319204399e-06,
+      "loss": 0.733,
+      "step": 5391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4985675491539114,
+      "learning_rate": 8.91824513765249e-06,
+      "loss": 0.7086,
+      "step": 5392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5089126612633614,
+      "learning_rate": 8.914970073508152e-06,
+      "loss": 0.7444,
+      "step": 5393
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.628177603423058,
+      "learning_rate": 8.91169512712685e-06,
+      "loss": 0.7051,
+      "step": 5394
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5772919443785387,
+      "learning_rate": 8.908420298864018e-06,
+      "loss": 0.7317,
+      "step": 5395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4723112256257591,
+      "learning_rate": 8.905145589075098e-06,
+      "loss": 0.7227,
+      "step": 5396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5801448338705688,
+      "learning_rate": 8.901870998115495e-06,
+      "loss": 0.6773,
+      "step": 5397
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5860986082201733,
+      "learning_rate": 8.898596526340624e-06,
+      "loss": 0.7343,
+      "step": 5398
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6039832133572458,
+      "learning_rate": 8.895322174105882e-06,
+      "loss": 0.805,
+      "step": 5399
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6155249124855937,
+      "learning_rate": 8.892047941766636e-06,
+      "loss": 0.7111,
+      "step": 5400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6067471312551125,
+      "learning_rate": 8.888773829678265e-06,
+      "loss": 0.7244,
+      "step": 5401
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.619332799189476,
+      "learning_rate": 8.885499838196114e-06,
+      "loss": 0.765,
+      "step": 5402
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6699781838918835,
+      "learning_rate": 8.882225967675526e-06,
+      "loss": 0.7806,
+      "step": 5403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6180503220313573,
+      "learning_rate": 8.878952218471833e-06,
+      "loss": 0.7772,
+      "step": 5404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7606649996624608,
+      "learning_rate": 8.87567859094034e-06,
+      "loss": 0.7342,
+      "step": 5405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6189646431863196,
+      "learning_rate": 8.872405085436356e-06,
+      "loss": 0.6012,
+      "step": 5406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6365844250183907,
+      "learning_rate": 8.869131702315169e-06,
+      "loss": 0.7184,
+      "step": 5407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5492476290084674,
+      "learning_rate": 8.865858441932047e-06,
+      "loss": 0.7093,
+      "step": 5408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5731924035028098,
+      "learning_rate": 8.862585304642257e-06,
+      "loss": 0.7088,
+      "step": 5409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5326160778305695,
+      "learning_rate": 8.859312290801044e-06,
+      "loss": 0.6615,
+      "step": 5410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.685641814822192,
+      "learning_rate": 8.856039400763644e-06,
+      "loss": 0.6843,
+      "step": 5411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6687141835760049,
+      "learning_rate": 8.852766634885277e-06,
+      "loss": 0.7037,
+      "step": 5412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5741605632580782,
+      "learning_rate": 8.849493993521153e-06,
+      "loss": 0.6725,
+      "step": 5413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5848257648940565,
+      "learning_rate": 8.846221477026463e-06,
+      "loss": 0.7393,
+      "step": 5414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4946988393052152,
+      "learning_rate": 8.842949085756389e-06,
+      "loss": 0.7276,
+      "step": 5415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.633398789443463,
+      "learning_rate": 8.839676820066095e-06,
+      "loss": 0.7075,
+      "step": 5416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5064528057849553,
+      "learning_rate": 8.836404680310739e-06,
+      "loss": 0.7695,
+      "step": 5417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6169451224176257,
+      "learning_rate": 8.833132666845459e-06,
+      "loss": 0.7448,
+      "step": 5418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5249218319990228,
+      "learning_rate": 8.82986078002538e-06,
+      "loss": 0.6317,
+      "step": 5419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6626818643466355,
+      "learning_rate": 8.826589020205611e-06,
+      "loss": 0.6781,
+      "step": 5420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5977691718018296,
+      "learning_rate": 8.823317387741258e-06,
+      "loss": 0.7576,
+      "step": 5421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5462585411188319,
+      "learning_rate": 8.8200458829874e-06,
+      "loss": 0.8659,
+      "step": 5422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5190555713203076,
+      "learning_rate": 8.816774506299106e-06,
+      "loss": 0.6913,
+      "step": 5423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6269163363674695,
+      "learning_rate": 8.813503258031443e-06,
+      "loss": 0.7462,
+      "step": 5424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.882660965495172,
+      "learning_rate": 8.81023213853944e-06,
+      "loss": 0.8512,
+      "step": 5425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5943514236393186,
+      "learning_rate": 8.806961148178133e-06,
+      "loss": 0.678,
+      "step": 5426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5669447477330756,
+      "learning_rate": 8.803690287302542e-06,
+      "loss": 0.668,
+      "step": 5427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6535602382835375,
+      "learning_rate": 8.800419556267655e-06,
+      "loss": 0.7596,
+      "step": 5428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6308465952597497,
+      "learning_rate": 8.797148955428467e-06,
+      "loss": 0.7934,
+      "step": 5429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.408309632362028,
+      "learning_rate": 8.793878485139954e-06,
+      "loss": 0.6747,
+      "step": 5430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6085099512031618,
+      "learning_rate": 8.790608145757066e-06,
+      "loss": 0.5726,
+      "step": 5431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5522606137544397,
+      "learning_rate": 8.787337937634755e-06,
+      "loss": 0.7439,
+      "step": 5432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5890313434061458,
+      "learning_rate": 8.784067861127942e-06,
+      "loss": 0.7395,
+      "step": 5433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.540243937101572,
+      "learning_rate": 8.780797916591548e-06,
+      "loss": 0.708,
+      "step": 5434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6690409088302076,
+      "learning_rate": 8.777528104380478e-06,
+      "loss": 0.7222,
+      "step": 5435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5706857558719922,
+      "learning_rate": 8.77425842484961e-06,
+      "loss": 0.786,
+      "step": 5436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4738961397397348,
+      "learning_rate": 8.770988878353827e-06,
+      "loss": 0.6863,
+      "step": 5437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.49909854563047,
+      "learning_rate": 8.767719465247979e-06,
+      "loss": 0.6873,
+      "step": 5438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7050766612344421,
+      "learning_rate": 8.76445018588691e-06,
+      "loss": 0.7244,
+      "step": 5439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5630758956722233,
+      "learning_rate": 8.761181040625457e-06,
+      "loss": 0.6779,
+      "step": 5440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5859976116499224,
+      "learning_rate": 8.757912029818424e-06,
+      "loss": 0.7304,
+      "step": 5441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6690118530119127,
+      "learning_rate": 8.75464315382062e-06,
+      "loss": 0.6904,
+      "step": 5442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.576408966950522,
+      "learning_rate": 8.751374412986822e-06,
+      "loss": 0.8189,
+      "step": 5443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6019837359423332,
+      "learning_rate": 8.748105807671806e-06,
+      "loss": 0.7219,
+      "step": 5444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8202596783529816,
+      "learning_rate": 8.744837338230331e-06,
+      "loss": 0.8334,
+      "step": 5445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5460218928791967,
+      "learning_rate": 8.74156900501713e-06,
+      "loss": 0.8658,
+      "step": 5446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6092606350579812,
+      "learning_rate": 8.738300808386934e-06,
+      "loss": 0.7317,
+      "step": 5447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6053711156944777,
+      "learning_rate": 8.73503274869446e-06,
+      "loss": 0.6881,
+      "step": 5448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6755703755220746,
+      "learning_rate": 8.731764826294394e-06,
+      "loss": 0.7608,
+      "step": 5449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.61294871243576,
+      "learning_rate": 8.728497041541426e-06,
+      "loss": 0.6088,
+      "step": 5450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4898392329846786,
+      "learning_rate": 8.725229394790217e-06,
+      "loss": 0.6994,
+      "step": 5451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5406794279721598,
+      "learning_rate": 8.721961886395423e-06,
+      "loss": 0.6649,
+      "step": 5452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5276423066659235,
+      "learning_rate": 8.718694516711684e-06,
+      "loss": 0.7596,
+      "step": 5453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.555067050858641,
+      "learning_rate": 8.715427286093611e-06,
+      "loss": 0.6365,
+      "step": 5454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4073917884142084,
+      "learning_rate": 8.712160194895825e-06,
+      "loss": 0.5868,
+      "step": 5455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5977461991934643,
+      "learning_rate": 8.708893243472905e-06,
+      "loss": 0.6963,
+      "step": 5456
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6408899699791568,
+      "learning_rate": 8.705626432179432e-06,
+      "loss": 0.6912,
+      "step": 5457
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5462815026322818,
+      "learning_rate": 8.702359761369972e-06,
+      "loss": 0.7436,
+      "step": 5458
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5104521935408541,
+      "learning_rate": 8.699093231399064e-06,
+      "loss": 0.726,
+      "step": 5459
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5145792617441518,
+      "learning_rate": 8.695826842621244e-06,
+      "loss": 0.6668,
+      "step": 5460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5567167872130345,
+      "learning_rate": 8.692560595391023e-06,
+      "loss": 0.6974,
+      "step": 5461
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.535631369442896,
+      "learning_rate": 8.689294490062906e-06,
+      "loss": 0.7519,
+      "step": 5462
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.525000280601043,
+      "learning_rate": 8.686028526991373e-06,
+      "loss": 0.6805,
+      "step": 5463
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3676512965755294,
+      "learning_rate": 8.682762706530897e-06,
+      "loss": 0.6684,
+      "step": 5464
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4824849114578038,
+      "learning_rate": 8.679497029035926e-06,
+      "loss": 0.7436,
+      "step": 5465
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6201369875081661,
+      "learning_rate": 8.676231494860907e-06,
+      "loss": 0.8019,
+      "step": 5466
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5794700859428037,
+      "learning_rate": 8.672966104360255e-06,
+      "loss": 0.6173,
+      "step": 5467
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.588515557807471,
+      "learning_rate": 8.669700857888382e-06,
+      "loss": 0.7549,
+      "step": 5468
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4362244278634058,
+      "learning_rate": 8.666435755799675e-06,
+      "loss": 0.6177,
+      "step": 5469
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4553152841062662,
+      "learning_rate": 8.663170798448511e-06,
+      "loss": 0.886,
+      "step": 5470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5276721104428754,
+      "learning_rate": 8.659905986189254e-06,
+      "loss": 0.6391,
+      "step": 5471
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3626636058873272,
+      "learning_rate": 8.656641319376244e-06,
+      "loss": 0.6458,
+      "step": 5472
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7623374037769408,
+      "learning_rate": 8.653376798363813e-06,
+      "loss": 0.7789,
+      "step": 5473
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.358068460640926,
+      "learning_rate": 8.650112423506268e-06,
+      "loss": 0.6103,
+      "step": 5474
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6080385386359963,
+      "learning_rate": 8.646848195157914e-06,
+      "loss": 0.7079,
+      "step": 5475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4797588538693942,
+      "learning_rate": 8.643584113673025e-06,
+      "loss": 0.5676,
+      "step": 5476
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5675290842063057,
+      "learning_rate": 8.640320179405867e-06,
+      "loss": 0.6131,
+      "step": 5477
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5496226897761918,
+      "learning_rate": 8.637056392710695e-06,
+      "loss": 0.5674,
+      "step": 5478
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5906722506085222,
+      "learning_rate": 8.633792753941733e-06,
+      "loss": 0.7608,
+      "step": 5479
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5845995315266523,
+      "learning_rate": 8.630529263453202e-06,
+      "loss": 0.8818,
+      "step": 5480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5647801024812311,
+      "learning_rate": 8.62726592159931e-06,
+      "loss": 0.7297,
+      "step": 5481
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5349042911362298,
+      "learning_rate": 8.624002728734228e-06,
+      "loss": 0.6783,
+      "step": 5482
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5358799651119135,
+      "learning_rate": 8.620739685212133e-06,
+      "loss": 0.6789,
+      "step": 5483
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6279261350508096,
+      "learning_rate": 8.61747679138718e-06,
+      "loss": 0.8028,
+      "step": 5484
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3911315916545854,
+      "learning_rate": 8.614214047613497e-06,
+      "loss": 0.6441,
+      "step": 5485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.638391124770247,
+      "learning_rate": 8.610951454245211e-06,
+      "loss": 0.715,
+      "step": 5486
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4050757193555106,
+      "learning_rate": 8.60768901163642e-06,
+      "loss": 0.6294,
+      "step": 5487
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4292014571964684,
+      "learning_rate": 8.604426720141211e-06,
+      "loss": 0.6052,
+      "step": 5488
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5718555849430869,
+      "learning_rate": 8.601164580113663e-06,
+      "loss": 0.6714,
+      "step": 5489
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6533540016966977,
+      "learning_rate": 8.597902591907821e-06,
+      "loss": 0.7301,
+      "step": 5490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7720517033439591,
+      "learning_rate": 8.594640755877728e-06,
+      "loss": 0.7721,
+      "step": 5491
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5961720840389233,
+      "learning_rate": 8.591379072377401e-06,
+      "loss": 0.6951,
+      "step": 5492
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8415406374214263,
+      "learning_rate": 8.588117541760848e-06,
+      "loss": 0.8032,
+      "step": 5493
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3633959194012253,
+      "learning_rate": 8.58485616438206e-06,
+      "loss": 0.6743,
+      "step": 5494
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.512968423653738,
+      "learning_rate": 8.581594940595002e-06,
+      "loss": 0.7512,
+      "step": 5495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4542667920517232,
+      "learning_rate": 8.578333870753635e-06,
+      "loss": 0.7837,
+      "step": 5496
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6243586978939764,
+      "learning_rate": 8.575072955211888e-06,
+      "loss": 0.6827,
+      "step": 5497
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5343117176073966,
+      "learning_rate": 8.57181219432369e-06,
+      "loss": 0.6556,
+      "step": 5498
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5582736131421835,
+      "learning_rate": 8.568551588442949e-06,
+      "loss": 0.7319,
+      "step": 5499
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5461179064783763,
+      "learning_rate": 8.565291137923542e-06,
+      "loss": 0.7341,
+      "step": 5500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7157349604771854,
+      "learning_rate": 8.562030843119346e-06,
+      "loss": 0.7225,
+      "step": 5501
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7871742129365662,
+      "learning_rate": 8.55877070438422e-06,
+      "loss": 0.7229,
+      "step": 5502
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9183352542015495,
+      "learning_rate": 8.55551072207199e-06,
+      "loss": 0.7484,
+      "step": 5503
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.464447201539831,
+      "learning_rate": 8.552250896536488e-06,
+      "loss": 0.7543,
+      "step": 5504
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8243480915817782,
+      "learning_rate": 8.548991228131503e-06,
+      "loss": 0.8253,
+      "step": 5505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.601641158363978,
+      "learning_rate": 8.545731717210832e-06,
+      "loss": 0.6864,
+      "step": 5506
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6520512376146232,
+      "learning_rate": 8.542472364128244e-06,
+      "loss": 0.6393,
+      "step": 5507
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5117404534779368,
+      "learning_rate": 8.539213169237483e-06,
+      "loss": 0.7381,
+      "step": 5508
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5655502787624929,
+      "learning_rate": 8.535954132892295e-06,
+      "loss": 0.7102,
+      "step": 5509
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7211773204730685,
+      "learning_rate": 8.532695255446384e-06,
+      "loss": 0.7406,
+      "step": 5510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6313306762180304,
+      "learning_rate": 8.529436537253458e-06,
+      "loss": 0.7661,
+      "step": 5511
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6058096645199098,
+      "learning_rate": 8.526177978667202e-06,
+      "loss": 0.7611,
+      "step": 5512
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4762936540269178,
+      "learning_rate": 8.522919580041276e-06,
+      "loss": 0.8195,
+      "step": 5513
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.751430534818543,
+      "learning_rate": 8.519661341729333e-06,
+      "loss": 0.6863,
+      "step": 5514
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.462329357830292,
+      "learning_rate": 8.516403264084998e-06,
+      "loss": 0.681,
+      "step": 5515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4524260115250442,
+      "learning_rate": 8.51314534746189e-06,
+      "loss": 0.6933,
+      "step": 5516
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5756065194086915,
+      "learning_rate": 8.509887592213604e-06,
+      "loss": 0.6209,
+      "step": 5517
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7025368438615078,
+      "learning_rate": 8.506629998693714e-06,
+      "loss": 0.7609,
+      "step": 5518
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7032354266244643,
+      "learning_rate": 8.503372567255787e-06,
+      "loss": 0.7251,
+      "step": 5519
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6639787231587528,
+      "learning_rate": 8.500115298253361e-06,
+      "loss": 0.6176,
+      "step": 5520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4583955236227215,
+      "learning_rate": 8.496858192039962e-06,
+      "loss": 0.615,
+      "step": 5521
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6337365754120245,
+      "learning_rate": 8.493601248969103e-06,
+      "loss": 0.8058,
+      "step": 5522
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7051305347035186,
+      "learning_rate": 8.490344469394271e-06,
+      "loss": 0.7225,
+      "step": 5523
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5575367248237868,
+      "learning_rate": 8.487087853668935e-06,
+      "loss": 0.63,
+      "step": 5524
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6287292527008408,
+      "learning_rate": 8.483831402146559e-06,
+      "loss": 0.6938,
+      "step": 5525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.50483887604819,
+      "learning_rate": 8.48057511518057e-06,
+      "loss": 0.6772,
+      "step": 5526
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7309490692074987,
+      "learning_rate": 8.477318993124393e-06,
+      "loss": 0.8651,
+      "step": 5527
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4125409131381073,
+      "learning_rate": 8.474063036331425e-06,
+      "loss": 0.712,
+      "step": 5528
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5606438706805952,
+      "learning_rate": 8.470807245155053e-06,
+      "loss": 0.6616,
+      "step": 5529
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4651607308021533,
+      "learning_rate": 8.46755161994864e-06,
+      "loss": 0.722,
+      "step": 5530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5445416441462756,
+      "learning_rate": 8.464296161065533e-06,
+      "loss": 0.7322,
+      "step": 5531
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5176082048874708,
+      "learning_rate": 8.461040868859065e-06,
+      "loss": 0.628,
+      "step": 5532
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5574105642679101,
+      "learning_rate": 8.45778574368254e-06,
+      "loss": 0.7133,
+      "step": 5533
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5858107938600134,
+      "learning_rate": 8.454530785889256e-06,
+      "loss": 0.762,
+      "step": 5534
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6696786532923031,
+      "learning_rate": 8.45127599583249e-06,
+      "loss": 0.7083,
+      "step": 5535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8353554641884566,
+      "learning_rate": 8.448021373865493e-06,
+      "loss": 0.7876,
+      "step": 5536
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5643365391993624,
+      "learning_rate": 8.444766920341509e-06,
+      "loss": 0.7114,
+      "step": 5537
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6023590781916328,
+      "learning_rate": 8.441512635613749e-06,
+      "loss": 0.6952,
+      "step": 5538
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.731932561372792,
+      "learning_rate": 8.438258520035421e-06,
+      "loss": 0.7214,
+      "step": 5539
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5965376353785534,
+      "learning_rate": 8.435004573959714e-06,
+      "loss": 0.6458,
+      "step": 5540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8887697894262163,
+      "learning_rate": 8.43175079773978e-06,
+      "loss": 0.7957,
+      "step": 5541
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5778138226547658,
+      "learning_rate": 8.428497191728773e-06,
+      "loss": 0.6641,
+      "step": 5542
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5779590248921962,
+      "learning_rate": 8.425243756279824e-06,
+      "loss": 0.7809,
+      "step": 5543
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5315990931760999,
+      "learning_rate": 8.421990491746036e-06,
+      "loss": 0.6848,
+      "step": 5544
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5889701751512135,
+      "learning_rate": 8.418737398480505e-06,
+      "loss": 0.6625,
+      "step": 5545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6874965203363166,
+      "learning_rate": 8.415484476836299e-06,
+      "loss": 0.7285,
+      "step": 5546
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.540099537700872,
+      "learning_rate": 8.412231727166473e-06,
+      "loss": 0.7185,
+      "step": 5547
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5670593760585358,
+      "learning_rate": 8.408979149824066e-06,
+      "loss": 0.749,
+      "step": 5548
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5576932693594387,
+      "learning_rate": 8.40572674516209e-06,
+      "loss": 0.5929,
+      "step": 5549
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4934320775822312,
+      "learning_rate": 8.402474513533547e-06,
+      "loss": 0.7322,
+      "step": 5550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6562092106422215,
+      "learning_rate": 8.39922245529141e-06,
+      "loss": 0.6973,
+      "step": 5551
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6473263063260772,
+      "learning_rate": 8.395970570788639e-06,
+      "loss": 0.6905,
+      "step": 5552
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.585602680375613,
+      "learning_rate": 8.392718860378184e-06,
+      "loss": 0.7053,
+      "step": 5553
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6454460458345783,
+      "learning_rate": 8.38946732441296e-06,
+      "loss": 0.7143,
+      "step": 5554
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4922249663072742,
+      "learning_rate": 8.386215963245873e-06,
+      "loss": 0.6574,
+      "step": 5555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5801772788488986,
+      "learning_rate": 8.382964777229805e-06,
+      "loss": 0.7476,
+      "step": 5556
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5681557434886617,
+      "learning_rate": 8.379713766717621e-06,
+      "loss": 0.6844,
+      "step": 5557
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6810740575671264,
+      "learning_rate": 8.376462932062173e-06,
+      "loss": 0.6406,
+      "step": 5558
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5934027970623843,
+      "learning_rate": 8.373212273616281e-06,
+      "loss": 0.7524,
+      "step": 5559
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6566698688257904,
+      "learning_rate": 8.369961791732758e-06,
+      "loss": 0.6743,
+      "step": 5560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6857676438146336,
+      "learning_rate": 8.366711486764392e-06,
+      "loss": 0.6712,
+      "step": 5561
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.705561999111611,
+      "learning_rate": 8.363461359063952e-06,
+      "loss": 0.6882,
+      "step": 5562
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.374891944612468,
+      "learning_rate": 8.360211408984192e-06,
+      "loss": 0.5937,
+      "step": 5563
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.574393404851258,
+      "learning_rate": 8.356961636877834e-06,
+      "loss": 0.7417,
+      "step": 5564
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7885837197042456,
+      "learning_rate": 8.353712043097598e-06,
+      "loss": 0.7578,
+      "step": 5565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6466916135457996,
+      "learning_rate": 8.350462627996177e-06,
+      "loss": 0.6979,
+      "step": 5566
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.655448490479182,
+      "learning_rate": 8.34721339192624e-06,
+      "loss": 0.7016,
+      "step": 5567
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6885302646033078,
+      "learning_rate": 8.343964335240443e-06,
+      "loss": 0.6853,
+      "step": 5568
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6656985984382529,
+      "learning_rate": 8.340715458291422e-06,
+      "loss": 0.6549,
+      "step": 5569
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5822177752381685,
+      "learning_rate": 8.337466761431785e-06,
+      "loss": 0.7849,
+      "step": 5570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.584122032281881,
+      "learning_rate": 8.334218245014138e-06,
+      "loss": 0.772,
+      "step": 5571
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.655308408893074,
+      "learning_rate": 8.330969909391046e-06,
+      "loss": 0.7676,
+      "step": 5572
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5107759446204707,
+      "learning_rate": 8.32772175491507e-06,
+      "loss": 0.7032,
+      "step": 5573
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5337381993316719,
+      "learning_rate": 8.324473781938749e-06,
+      "loss": 0.743,
+      "step": 5574
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6892896481283168,
+      "learning_rate": 8.321225990814592e-06,
+      "loss": 0.6367,
+      "step": 5575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5491307737483462,
+      "learning_rate": 8.317978381895105e-06,
+      "loss": 0.7983,
+      "step": 5576
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6842601260701693,
+      "learning_rate": 8.314730955532757e-06,
+      "loss": 0.7657,
+      "step": 5577
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4904895744526254,
+      "learning_rate": 8.311483712080011e-06,
+      "loss": 0.6465,
+      "step": 5578
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5253575724269708,
+      "learning_rate": 8.308236651889304e-06,
+      "loss": 0.5721,
+      "step": 5579
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6042099863240251,
+      "learning_rate": 8.30498977531305e-06,
+      "loss": 0.7276,
+      "step": 5580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.497215582210722,
+      "learning_rate": 8.301743082703651e-06,
+      "loss": 0.6767,
+      "step": 5581
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5552740736267856,
+      "learning_rate": 8.29849657441348e-06,
+      "loss": 0.7603,
+      "step": 5582
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5397565487267744,
+      "learning_rate": 8.2952502507949e-06,
+      "loss": 0.6586,
+      "step": 5583
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6627728598845204,
+      "learning_rate": 8.292004112200245e-06,
+      "loss": 0.6835,
+      "step": 5584
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6729356422110861,
+      "learning_rate": 8.28875815898183e-06,
+      "loss": 0.7202,
+      "step": 5585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5953255361916052,
+      "learning_rate": 8.285512391491964e-06,
+      "loss": 0.7248,
+      "step": 5586
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5983069282455369,
+      "learning_rate": 8.28226681008291e-06,
+      "loss": 0.6544,
+      "step": 5587
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3969342118400652,
+      "learning_rate": 8.27902141510693e-06,
+      "loss": 0.5857,
+      "step": 5588
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6125249307841638,
+      "learning_rate": 8.275776206916268e-06,
+      "loss": 0.6612,
+      "step": 5589
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6681447234561344,
+      "learning_rate": 8.27253118586313e-06,
+      "loss": 0.7215,
+      "step": 5590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5914349153850604,
+      "learning_rate": 8.269286352299723e-06,
+      "loss": 0.6635,
+      "step": 5591
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7033247822836768,
+      "learning_rate": 8.266041706578212e-06,
+      "loss": 0.6536,
+      "step": 5592
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6737519134798131,
+      "learning_rate": 8.262797249050758e-06,
+      "loss": 0.7259,
+      "step": 5593
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6019239391867346,
+      "learning_rate": 8.259552980069499e-06,
+      "loss": 0.7038,
+      "step": 5594
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7608015334026599,
+      "learning_rate": 8.256308899986543e-06,
+      "loss": 0.5839,
+      "step": 5595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6677107873015877,
+      "learning_rate": 8.253065009153988e-06,
+      "loss": 0.6422,
+      "step": 5596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5685334631977799,
+      "learning_rate": 8.249821307923911e-06,
+      "loss": 0.7272,
+      "step": 5597
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4941559661351542,
+      "learning_rate": 8.246577796648357e-06,
+      "loss": 0.7019,
+      "step": 5598
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6341561834848046,
+      "learning_rate": 8.243334475679367e-06,
+      "loss": 0.7615,
+      "step": 5599
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7031917858396979,
+      "learning_rate": 8.240091345368944e-06,
+      "loss": 0.8472,
+      "step": 5600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4470246442660457,
+      "learning_rate": 8.236848406069084e-06,
+      "loss": 0.6536,
+      "step": 5601
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7029488235864563,
+      "learning_rate": 8.23360565813176e-06,
+      "loss": 0.7819,
+      "step": 5602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7033324038098736,
+      "learning_rate": 8.230363101908916e-06,
+      "loss": 0.7275,
+      "step": 5603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.485534242872215,
+      "learning_rate": 8.227120737752486e-06,
+      "loss": 0.6535,
+      "step": 5604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6491114795083066,
+      "learning_rate": 8.223878566014372e-06,
+      "loss": 0.7522,
+      "step": 5605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.688704750751506,
+      "learning_rate": 8.220636587046464e-06,
+      "loss": 0.763,
+      "step": 5606
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3452689780904759,
+      "learning_rate": 8.217394801200632e-06,
+      "loss": 0.6435,
+      "step": 5607
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6312682950472936,
+      "learning_rate": 8.214153208828714e-06,
+      "loss": 0.7451,
+      "step": 5608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5168308517665703,
+      "learning_rate": 8.210911810282542e-06,
+      "loss": 0.6365,
+      "step": 5609
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6131248031193193,
+      "learning_rate": 8.20767060591391e-06,
+      "loss": 0.6881,
+      "step": 5610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5979577897434138,
+      "learning_rate": 8.204429596074605e-06,
+      "loss": 0.7088,
+      "step": 5611
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4937526190944859,
+      "learning_rate": 8.201188781116392e-06,
+      "loss": 0.6796,
+      "step": 5612
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.45679874555144,
+      "learning_rate": 8.197948161391003e-06,
+      "loss": 0.7388,
+      "step": 5613
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5562185392172332,
+      "learning_rate": 8.194707737250158e-06,
+      "loss": 0.7071,
+      "step": 5614
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7053520652444238,
+      "learning_rate": 8.191467509045564e-06,
+      "loss": 0.7217,
+      "step": 5615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5433041988831695,
+      "learning_rate": 8.188227477128883e-06,
+      "loss": 0.7122,
+      "step": 5616
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5797508292402456,
+      "learning_rate": 8.18498764185178e-06,
+      "loss": 0.6357,
+      "step": 5617
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.47683890933061,
+      "learning_rate": 8.181748003565883e-06,
+      "loss": 0.6484,
+      "step": 5618
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5348046552396966,
+      "learning_rate": 8.178508562622804e-06,
+      "loss": 0.7089,
+      "step": 5619
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.414162890397765,
+      "learning_rate": 8.175269319374142e-06,
+      "loss": 0.7159,
+      "step": 5620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7791038023349213,
+      "learning_rate": 8.172030274171452e-06,
+      "loss": 0.7431,
+      "step": 5621
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4784290773633204,
+      "learning_rate": 8.168791427366297e-06,
+      "loss": 0.7103,
+      "step": 5622
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.635880216665212,
+      "learning_rate": 8.16555277931019e-06,
+      "loss": 0.7444,
+      "step": 5623
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7363639324473743,
+      "learning_rate": 8.162314330354642e-06,
+      "loss": 0.733,
+      "step": 5624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7253204538491358,
+      "learning_rate": 8.159076080851139e-06,
+      "loss": 0.7552,
+      "step": 5625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5681074189686877,
+      "learning_rate": 8.155838031151134e-06,
+      "loss": 0.7665,
+      "step": 5626
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5958043823114054,
+      "learning_rate": 8.152600181606073e-06,
+      "loss": 0.6868,
+      "step": 5627
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5497556316778451,
+      "learning_rate": 8.149362532567374e-06,
+      "loss": 0.6495,
+      "step": 5628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5626805277983316,
+      "learning_rate": 8.146125084386428e-06,
+      "loss": 0.6569,
+      "step": 5629
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6780030139091988,
+      "learning_rate": 8.142887837414614e-06,
+      "loss": 0.8537,
+      "step": 5630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5189408676389409,
+      "learning_rate": 8.139650792003286e-06,
+      "loss": 0.8312,
+      "step": 5631
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4776069199360429,
+      "learning_rate": 8.136413948503767e-06,
+      "loss": 0.6339,
+      "step": 5632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5138991319169117,
+      "learning_rate": 8.133177307267376e-06,
+      "loss": 0.7134,
+      "step": 5633
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5731729926117684,
+      "learning_rate": 8.12994086864539e-06,
+      "loss": 0.7459,
+      "step": 5634
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5803257301451343,
+      "learning_rate": 8.126704632989082e-06,
+      "loss": 0.7218,
+      "step": 5635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6209290479538236,
+      "learning_rate": 8.123468600649686e-06,
+      "loss": 0.7538,
+      "step": 5636
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4808690714988408,
+      "learning_rate": 8.120232771978432e-06,
+      "loss": 0.689,
+      "step": 5637
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.676758702387847,
+      "learning_rate": 8.116997147326515e-06,
+      "loss": 0.7705,
+      "step": 5638
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5484496099134843,
+      "learning_rate": 8.113761727045106e-06,
+      "loss": 0.737,
+      "step": 5639
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7164143999551997,
+      "learning_rate": 8.110526511485371e-06,
+      "loss": 0.7558,
+      "step": 5640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5754148116572044,
+      "learning_rate": 8.10729150099843e-06,
+      "loss": 0.6998,
+      "step": 5641
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7140088046757154,
+      "learning_rate": 8.104056695935396e-06,
+      "loss": 0.7623,
+      "step": 5642
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7869442676574774,
+      "learning_rate": 8.100822096647365e-06,
+      "loss": 0.7454,
+      "step": 5643
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4764597743231278,
+      "learning_rate": 8.09758770348539e-06,
+      "loss": 0.6513,
+      "step": 5644
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6279896509723633,
+      "learning_rate": 8.094353516800524e-06,
+      "loss": 0.6936,
+      "step": 5645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5207184246056769,
+      "learning_rate": 8.091119536943779e-06,
+      "loss": 0.7996,
+      "step": 5646
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0178330689034416,
+      "learning_rate": 8.087885764266156e-06,
+      "loss": 0.7302,
+      "step": 5647
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5985413437981397,
+      "learning_rate": 8.084652199118636e-06,
+      "loss": 0.7385,
+      "step": 5648
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6947296632777942,
+      "learning_rate": 8.081418841852163e-06,
+      "loss": 0.7599,
+      "step": 5649
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.485467514936313,
+      "learning_rate": 8.07818569281767e-06,
+      "loss": 0.7569,
+      "step": 5650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6350537395684603,
+      "learning_rate": 8.074952752366072e-06,
+      "loss": 0.6527,
+      "step": 5651
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4772057181632665,
+      "learning_rate": 8.071720020848245e-06,
+      "loss": 0.677,
+      "step": 5652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5288422750275847,
+      "learning_rate": 8.068487498615057e-06,
+      "loss": 0.6706,
+      "step": 5653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5773263458771583,
+      "learning_rate": 8.065255186017342e-06,
+      "loss": 0.7145,
+      "step": 5654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4698615244133397,
+      "learning_rate": 8.06202308340592e-06,
+      "loss": 0.6992,
+      "step": 5655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0458570117068953,
+      "learning_rate": 8.05879119113159e-06,
+      "loss": 0.7137,
+      "step": 5656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5528151714159797,
+      "learning_rate": 8.055559509545115e-06,
+      "loss": 0.5657,
+      "step": 5657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5618631597788561,
+      "learning_rate": 8.052328038997252e-06,
+      "loss": 0.7858,
+      "step": 5658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6302595225217231,
+      "learning_rate": 8.04909677983872e-06,
+      "loss": 0.6427,
+      "step": 5659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8190481089776078,
+      "learning_rate": 8.04586573242022e-06,
+      "loss": 0.803,
+      "step": 5660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6958660962032366,
+      "learning_rate": 8.042634897092443e-06,
+      "loss": 0.7755,
+      "step": 5661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.742053888634955,
+      "learning_rate": 8.039404274206031e-06,
+      "loss": 0.6488,
+      "step": 5662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5210478365225468,
+      "learning_rate": 8.03617386411163e-06,
+      "loss": 0.6826,
+      "step": 5663
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4114991514657393,
+      "learning_rate": 8.032943667159843e-06,
+      "loss": 0.6926,
+      "step": 5664
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4134598644524605,
+      "learning_rate": 8.029713683701259e-06,
+      "loss": 0.6353,
+      "step": 5665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5362933596085961,
+      "learning_rate": 8.026483914086446e-06,
+      "loss": 0.7266,
+      "step": 5666
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4561356510584063,
+      "learning_rate": 8.023254358665938e-06,
+      "loss": 0.7336,
+      "step": 5667
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5192674346006763,
+      "learning_rate": 8.020025017790261e-06,
+      "loss": 0.7731,
+      "step": 5668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5640800467790366,
+      "learning_rate": 8.0167958918099e-06,
+      "loss": 0.716,
+      "step": 5669
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.456511074126383,
+      "learning_rate": 8.013566981075331e-06,
+      "loss": 0.6562,
+      "step": 5670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6136566932546004,
+      "learning_rate": 8.010338285937006e-06,
+      "loss": 0.7981,
+      "step": 5671
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.669060168683187,
+      "learning_rate": 8.007109806745341e-06,
+      "loss": 0.7027,
+      "step": 5672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4881521535751443,
+      "learning_rate": 8.003881543850742e-06,
+      "loss": 0.7885,
+      "step": 5673
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5634498701023354,
+      "learning_rate": 8.00065349760359e-06,
+      "loss": 0.6158,
+      "step": 5674
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6285482943204677,
+      "learning_rate": 7.99742566835423e-06,
+      "loss": 0.7625,
+      "step": 5675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5147804177704658,
+      "learning_rate": 7.994198056453001e-06,
+      "loss": 0.7938,
+      "step": 5676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7054944824732357,
+      "learning_rate": 7.9909706622502e-06,
+      "loss": 0.8145,
+      "step": 5677
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4443496050342668,
+      "learning_rate": 7.987743486096119e-06,
+      "loss": 0.7224,
+      "step": 5678
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7038128500780676,
+      "learning_rate": 7.984516528341016e-06,
+      "loss": 0.6913,
+      "step": 5679
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.582191849396049,
+      "learning_rate": 7.981289789335122e-06,
+      "loss": 0.7606,
+      "step": 5680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7291647907979886,
+      "learning_rate": 7.978063269428656e-06,
+      "loss": 0.7354,
+      "step": 5681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5886724144698394,
+      "learning_rate": 7.974836968971799e-06,
+      "loss": 0.7044,
+      "step": 5682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5544304168529097,
+      "learning_rate": 7.97161088831472e-06,
+      "loss": 0.7389,
+      "step": 5683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6719397258805486,
+      "learning_rate": 7.968385027807558e-06,
+      "loss": 0.7663,
+      "step": 5684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4880037468847649,
+      "learning_rate": 7.965159387800434e-06,
+      "loss": 0.6278,
+      "step": 5685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.602445909793838,
+      "learning_rate": 7.961933968643435e-06,
+      "loss": 0.8443,
+      "step": 5686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7395603370748483,
+      "learning_rate": 7.95870877068663e-06,
+      "loss": 0.7678,
+      "step": 5687
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8056856860517443,
+      "learning_rate": 7.955483794280068e-06,
+      "loss": 0.6395,
+      "step": 5688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.55847808067052,
+      "learning_rate": 7.952259039773767e-06,
+      "loss": 0.6431,
+      "step": 5689
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.687208438007986,
+      "learning_rate": 7.949034507517721e-06,
+      "loss": 0.8264,
+      "step": 5690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.835421354444477,
+      "learning_rate": 7.94581019786191e-06,
+      "loss": 0.7163,
+      "step": 5691
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7973409383727332,
+      "learning_rate": 7.942586111156277e-06,
+      "loss": 0.7155,
+      "step": 5692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7173157692995349,
+      "learning_rate": 7.939362247750745e-06,
+      "loss": 0.7497,
+      "step": 5693
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.675781627264294,
+      "learning_rate": 7.93613860799522e-06,
+      "loss": 0.7794,
+      "step": 5694
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5541808483981965,
+      "learning_rate": 7.932915192239571e-06,
+      "loss": 0.6026,
+      "step": 5695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5952182515805586,
+      "learning_rate": 7.929692000833653e-06,
+      "loss": 0.6893,
+      "step": 5696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6961654663081456,
+      "learning_rate": 7.926469034127292e-06,
+      "loss": 0.7462,
+      "step": 5697
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.732500632150684,
+      "learning_rate": 7.923246292470292e-06,
+      "loss": 0.7884,
+      "step": 5698
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6452456021606285,
+      "learning_rate": 7.920023776212433e-06,
+      "loss": 0.7367,
+      "step": 5699
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.52846216501172,
+      "learning_rate": 7.91680148570346e-06,
+      "loss": 0.6121,
+      "step": 5700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6674960297503236,
+      "learning_rate": 7.91357942129311e-06,
+      "loss": 0.7909,
+      "step": 5701
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5036535214427527,
+      "learning_rate": 7.910357583331088e-06,
+      "loss": 0.7316,
+      "step": 5702
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.652839330197833,
+      "learning_rate": 7.90713597216707e-06,
+      "loss": 0.7686,
+      "step": 5703
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4637652515969,
+      "learning_rate": 7.903914588150716e-06,
+      "loss": 0.739,
+      "step": 5704
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6957518798495284,
+      "learning_rate": 7.90069343163165e-06,
+      "loss": 0.7086,
+      "step": 5705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6581052593969878,
+      "learning_rate": 7.897472502959484e-06,
+      "loss": 0.6988,
+      "step": 5706
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.716293328997258,
+      "learning_rate": 7.894251802483803e-06,
+      "loss": 0.7448,
+      "step": 5707
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.608428356318047,
+      "learning_rate": 7.891031330554151e-06,
+      "loss": 0.7065,
+      "step": 5708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4342559198109504,
+      "learning_rate": 7.88781108752007e-06,
+      "loss": 0.7052,
+      "step": 5709
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8107927017895886,
+      "learning_rate": 7.884591073731068e-06,
+      "loss": 0.7711,
+      "step": 5710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5861194643001888,
+      "learning_rate": 7.88137128953662e-06,
+      "loss": 0.7277,
+      "step": 5711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.609787288473546,
+      "learning_rate": 7.87815173528619e-06,
+      "loss": 0.6833,
+      "step": 5712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.648058051240888,
+      "learning_rate": 7.874932411329203e-06,
+      "loss": 0.6429,
+      "step": 5713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7289025911897014,
+      "learning_rate": 7.871713318015071e-06,
+      "loss": 0.6381,
+      "step": 5714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7201473743909554,
+      "learning_rate": 7.868494455693181e-06,
+      "loss": 0.7717,
+      "step": 5715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5865041465364405,
+      "learning_rate": 7.86527582471288e-06,
+      "loss": 0.653,
+      "step": 5716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.577293936850371,
+      "learning_rate": 7.862057425423509e-06,
+      "loss": 0.753,
+      "step": 5717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.414651766000381,
+      "learning_rate": 7.858839258174365e-06,
+      "loss": 0.6993,
+      "step": 5718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6155072297570676,
+      "learning_rate": 7.855621323314736e-06,
+      "loss": 0.7568,
+      "step": 5719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.575719982898642,
+      "learning_rate": 7.852403621193883e-06,
+      "loss": 0.7807,
+      "step": 5720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6118080727936386,
+      "learning_rate": 7.849186152161028e-06,
+      "loss": 0.7623,
+      "step": 5721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6776561690628125,
+      "learning_rate": 7.845968916565385e-06,
+      "loss": 0.6105,
+      "step": 5722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6816273038027498,
+      "learning_rate": 7.842751914756128e-06,
+      "loss": 0.7343,
+      "step": 5723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.521521598263868,
+      "learning_rate": 7.839535147082414e-06,
+      "loss": 0.7389,
+      "step": 5724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6662923031868875,
+      "learning_rate": 7.836318613893377e-06,
+      "loss": 0.7188,
+      "step": 5725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5968447870884663,
+      "learning_rate": 7.833102315538116e-06,
+      "loss": 0.6968,
+      "step": 5726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.620655229813537,
+      "learning_rate": 7.829886252365712e-06,
+      "loss": 0.7694,
+      "step": 5727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5244934459907622,
+      "learning_rate": 7.82667042472522e-06,
+      "loss": 0.658,
+      "step": 5728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6190112998576944,
+      "learning_rate": 7.823454832965666e-06,
+      "loss": 0.706,
+      "step": 5729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7098127138691492,
+      "learning_rate": 7.820239477436055e-06,
+      "loss": 0.7253,
+      "step": 5730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6048506977406953,
+      "learning_rate": 7.817024358485357e-06,
+      "loss": 0.7531,
+      "step": 5731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5235882405052927,
+      "learning_rate": 7.813809476462525e-06,
+      "loss": 0.6348,
+      "step": 5732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4798292434357883,
+      "learning_rate": 7.810594831716494e-06,
+      "loss": 0.5956,
+      "step": 5733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5262332112480077,
+      "learning_rate": 7.807380424596149e-06,
+      "loss": 0.7542,
+      "step": 5734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.442225125841356,
+      "learning_rate": 7.804166255450372e-06,
+      "loss": 0.7161,
+      "step": 5735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.641892390278661,
+      "learning_rate": 7.80095232462801e-06,
+      "loss": 0.851,
+      "step": 5736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.620114405413643,
+      "learning_rate": 7.797738632477881e-06,
+      "loss": 0.7402,
+      "step": 5737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4693871991300977,
+      "learning_rate": 7.794525179348786e-06,
+      "loss": 0.6569,
+      "step": 5738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5528409369558045,
+      "learning_rate": 7.791311965589494e-06,
+      "loss": 0.6966,
+      "step": 5739
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5940979057868363,
+      "learning_rate": 7.788098991548749e-06,
+      "loss": 0.7308,
+      "step": 5740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5244751204342553,
+      "learning_rate": 7.784886257575264e-06,
+      "loss": 0.6187,
+      "step": 5741
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6675693980486168,
+      "learning_rate": 7.781673764017738e-06,
+      "loss": 0.6508,
+      "step": 5742
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.624044453668377,
+      "learning_rate": 7.778461511224835e-06,
+      "loss": 0.6836,
+      "step": 5743
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8273438127319237,
+      "learning_rate": 7.775249499545189e-06,
+      "loss": 0.6414,
+      "step": 5744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8859707954802163,
+      "learning_rate": 7.772037729327423e-06,
+      "loss": 0.7952,
+      "step": 5745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6982793251008128,
+      "learning_rate": 7.76882620092012e-06,
+      "loss": 0.8613,
+      "step": 5746
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.410673546463811,
+      "learning_rate": 7.76561491467184e-06,
+      "loss": 0.7405,
+      "step": 5747
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4415440808988367,
+      "learning_rate": 7.762403870931123e-06,
+      "loss": 0.6729,
+      "step": 5748
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5811451596234793,
+      "learning_rate": 7.75919307004647e-06,
+      "loss": 0.7217,
+      "step": 5749
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6540154907977642,
+      "learning_rate": 7.755982512366366e-06,
+      "loss": 0.6948,
+      "step": 5750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6137901831730095,
+      "learning_rate": 7.75277219823927e-06,
+      "loss": 0.6641,
+      "step": 5751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6472949529054923,
+      "learning_rate": 7.749562128013606e-06,
+      "loss": 0.6697,
+      "step": 5752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6150286746139912,
+      "learning_rate": 7.746352302037786e-06,
+      "loss": 0.6422,
+      "step": 5753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.801711018578529,
+      "learning_rate": 7.743142720660175e-06,
+      "loss": 0.6853,
+      "step": 5754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4361518027545168,
+      "learning_rate": 7.73993338422913e-06,
+      "loss": 0.6891,
+      "step": 5755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7345840777511297,
+      "learning_rate": 7.736724293092975e-06,
+      "loss": 0.7067,
+      "step": 5756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.481089883684841,
+      "learning_rate": 7.733515447600001e-06,
+      "loss": 0.6426,
+      "step": 5757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4978401850787875,
+      "learning_rate": 7.730306848098485e-06,
+      "loss": 0.7194,
+      "step": 5758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5186636663429582,
+      "learning_rate": 7.727098494936663e-06,
+      "loss": 0.7443,
+      "step": 5759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.60008866516873,
+      "learning_rate": 7.723890388462755e-06,
+      "loss": 0.7472,
+      "step": 5760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4684241101960502,
+      "learning_rate": 7.720682529024954e-06,
+      "loss": 0.6613,
+      "step": 5761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3717717677756704,
+      "learning_rate": 7.717474916971417e-06,
+      "loss": 0.5526,
+      "step": 5762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4948051573294627,
+      "learning_rate": 7.714267552650283e-06,
+      "loss": 0.6751,
+      "step": 5763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6615283937090968,
+      "learning_rate": 7.711060436409666e-06,
+      "loss": 0.8354,
+      "step": 5764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.504730171070354,
+      "learning_rate": 7.707853568597638e-06,
+      "loss": 0.7561,
+      "step": 5765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7639588172269918,
+      "learning_rate": 7.704646949562264e-06,
+      "loss": 0.7053,
+      "step": 5766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6477621135946416,
+      "learning_rate": 7.701440579651566e-06,
+      "loss": 0.7342,
+      "step": 5767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4672356032530092,
+      "learning_rate": 7.698234459213545e-06,
+      "loss": 0.6801,
+      "step": 5768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6686897139751138,
+      "learning_rate": 7.695028588596185e-06,
+      "loss": 0.6892,
+      "step": 5769
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5214058904728014,
+      "learning_rate": 7.69182296814742e-06,
+      "loss": 0.6733,
+      "step": 5770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7885394268906438,
+      "learning_rate": 7.688617598215182e-06,
+      "loss": 0.7831,
+      "step": 5771
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6048047448421283,
+      "learning_rate": 7.685412479147352e-06,
+      "loss": 0.6657,
+      "step": 5772
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5551971775290863,
+      "learning_rate": 7.6822076112918e-06,
+      "loss": 0.6063,
+      "step": 5773
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6143685528810567,
+      "learning_rate": 7.679002994996372e-06,
+      "loss": 0.7048,
+      "step": 5774
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.657918462223705,
+      "learning_rate": 7.675798630608867e-06,
+      "loss": 0.771,
+      "step": 5775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.528457546476097,
+      "learning_rate": 7.672594518477078e-06,
+      "loss": 0.7309,
+      "step": 5776
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5279410934335003,
+      "learning_rate": 7.669390658948755e-06,
+      "loss": 0.7658,
+      "step": 5777
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6382782909336935,
+      "learning_rate": 7.66618705237163e-06,
+      "loss": 0.7689,
+      "step": 5778
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5531108260546986,
+      "learning_rate": 7.662983699093406e-06,
+      "loss": 0.7474,
+      "step": 5779
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5019858046942969,
+      "learning_rate": 7.65978059946175e-06,
+      "loss": 0.581,
+      "step": 5780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6333470851583256,
+      "learning_rate": 7.656577753824314e-06,
+      "loss": 0.7437,
+      "step": 5781
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6317538425852873,
+      "learning_rate": 7.65337516252872e-06,
+      "loss": 0.6981,
+      "step": 5782
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.65717576290963,
+      "learning_rate": 7.65017282592255e-06,
+      "loss": 0.6735,
+      "step": 5783
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6307719724360934,
+      "learning_rate": 7.64697074435338e-06,
+      "loss": 0.6793,
+      "step": 5784
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5764880566930657,
+      "learning_rate": 7.64376891816873e-06,
+      "loss": 0.707,
+      "step": 5785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5477253630397978,
+      "learning_rate": 7.64056734771612e-06,
+      "loss": 0.6924,
+      "step": 5786
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5882482658079655,
+      "learning_rate": 7.637366033343028e-06,
+      "loss": 0.6662,
+      "step": 5787
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4886743464528789,
+      "learning_rate": 7.634164975396903e-06,
+      "loss": 0.6774,
+      "step": 5788
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5966823968369421,
+      "learning_rate": 7.630964174225175e-06,
+      "loss": 0.732,
+      "step": 5789
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.47748735264666,
+      "learning_rate": 7.627763630175236e-06,
+      "loss": 0.7289,
+      "step": 5790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5900714815590495,
+      "learning_rate": 7.624563343594457e-06,
+      "loss": 0.6183,
+      "step": 5791
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4729519261324666,
+      "learning_rate": 7.621363314830182e-06,
+      "loss": 0.6789,
+      "step": 5792
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5556688222110906,
+      "learning_rate": 7.618163544229722e-06,
+      "loss": 0.6508,
+      "step": 5793
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.642197194926469,
+      "learning_rate": 7.614964032140359e-06,
+      "loss": 0.6123,
+      "step": 5794
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5312811606471664,
+      "learning_rate": 7.611764778909352e-06,
+      "loss": 0.7497,
+      "step": 5795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.612451680261775,
+      "learning_rate": 7.608565784883932e-06,
+      "loss": 0.7227,
+      "step": 5796
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.619756588082997,
+      "learning_rate": 7.6053670504112995e-06,
+      "loss": 0.7505,
+      "step": 5797
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5693701631566446,
+      "learning_rate": 7.602168575838622e-06,
+      "loss": 0.7009,
+      "step": 5798
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.856737086352245,
+      "learning_rate": 7.598970361513052e-06,
+      "loss": 0.7856,
+      "step": 5799
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5410044281344544,
+      "learning_rate": 7.5957724077816985e-06,
+      "loss": 0.7093,
+      "step": 5800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7184048220446808,
+      "learning_rate": 7.5925747149916515e-06,
+      "loss": 0.6581,
+      "step": 5801
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6669867702953898,
+      "learning_rate": 7.589377283489976e-06,
+      "loss": 0.7804,
+      "step": 5802
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6790229508054892,
+      "learning_rate": 7.586180113623694e-06,
+      "loss": 0.769,
+      "step": 5803
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4762824768291218,
+      "learning_rate": 7.582983205739815e-06,
+      "loss": 0.6036,
+      "step": 5804
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3934053469588565,
+      "learning_rate": 7.579786560185311e-06,
+      "loss": 0.6886,
+      "step": 5805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6255197560643635,
+      "learning_rate": 7.576590177307125e-06,
+      "loss": 0.6945,
+      "step": 5806
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6613780786906946,
+      "learning_rate": 7.573394057452181e-06,
+      "loss": 0.7167,
+      "step": 5807
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.570481880582874,
+      "learning_rate": 7.570198200967363e-06,
+      "loss": 0.7472,
+      "step": 5808
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.617755903243908,
+      "learning_rate": 7.5670026081995295e-06,
+      "loss": 0.7497,
+      "step": 5809
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7559979574300808,
+      "learning_rate": 7.563807279495521e-06,
+      "loss": 0.7635,
+      "step": 5810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7041260538063296,
+      "learning_rate": 7.560612215202129e-06,
+      "loss": 0.7424,
+      "step": 5811
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5673151751986434,
+      "learning_rate": 7.557417415666138e-06,
+      "loss": 0.7054,
+      "step": 5812
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.642617769709063,
+      "learning_rate": 7.554222881234284e-06,
+      "loss": 0.6868,
+      "step": 5813
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4400619885328076,
+      "learning_rate": 7.55102861225329e-06,
+      "loss": 0.7142,
+      "step": 5814
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.52967714882465,
+      "learning_rate": 7.547834609069846e-06,
+      "loss": 0.7085,
+      "step": 5815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4760264857506311,
+      "learning_rate": 7.544640872030604e-06,
+      "loss": 0.6837,
+      "step": 5816
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6786991582733812,
+      "learning_rate": 7.5414474014822e-06,
+      "loss": 0.8335,
+      "step": 5817
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8209395656572913,
+      "learning_rate": 7.538254197771231e-06,
+      "loss": 0.7985,
+      "step": 5818
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6453101136824253,
+      "learning_rate": 7.535061261244271e-06,
+      "loss": 0.7504,
+      "step": 5819
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5088877840540806,
+      "learning_rate": 7.5318685922478675e-06,
+      "loss": 0.6652,
+      "step": 5820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3748999206010954,
+      "learning_rate": 7.528676191128528e-06,
+      "loss": 0.7178,
+      "step": 5821
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.60091876485254,
+      "learning_rate": 7.525484058232739e-06,
+      "loss": 0.8099,
+      "step": 5822
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4644531630938904,
+      "learning_rate": 7.522292193906964e-06,
+      "loss": 0.6834,
+      "step": 5823
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.487917947088439,
+      "learning_rate": 7.5191005984976196e-06,
+      "loss": 0.6411,
+      "step": 5824
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6010186255910743,
+      "learning_rate": 7.515909272351112e-06,
+      "loss": 0.7479,
+      "step": 5825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6386631361679176,
+      "learning_rate": 7.512718215813802e-06,
+      "loss": 0.7278,
+      "step": 5826
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7203597276680547,
+      "learning_rate": 7.509527429232033e-06,
+      "loss": 0.7027,
+      "step": 5827
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7403021291942407,
+      "learning_rate": 7.506336912952121e-06,
+      "loss": 0.6961,
+      "step": 5828
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.649231460060806,
+      "learning_rate": 7.503146667320334e-06,
+      "loss": 0.6828,
+      "step": 5829
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.548994166079608,
+      "learning_rate": 7.499956692682935e-06,
+      "loss": 0.684,
+      "step": 5830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5992775479851562,
+      "learning_rate": 7.4967669893861364e-06,
+      "loss": 0.6688,
+      "step": 5831
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5744237446744,
+      "learning_rate": 7.493577557776135e-06,
+      "loss": 0.7714,
+      "step": 5832
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5185226876303048,
+      "learning_rate": 7.490388398199098e-06,
+      "loss": 0.7452,
+      "step": 5833
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8450325476321954,
+      "learning_rate": 7.487199511001148e-06,
+      "loss": 0.7273,
+      "step": 5834
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6248177114263815,
+      "learning_rate": 7.4840108965284016e-06,
+      "loss": 0.7375,
+      "step": 5835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7510103642438084,
+      "learning_rate": 7.480822555126923e-06,
+      "loss": 0.8054,
+      "step": 5836
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4450252781350619,
+      "learning_rate": 7.477634487142759e-06,
+      "loss": 0.7438,
+      "step": 5837
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5620357709252137,
+      "learning_rate": 7.474446692921931e-06,
+      "loss": 0.8239,
+      "step": 5838
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7434351076894188,
+      "learning_rate": 7.471259172810417e-06,
+      "loss": 0.7382,
+      "step": 5839
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7191352077450146,
+      "learning_rate": 7.468071927154173e-06,
+      "loss": 0.7162,
+      "step": 5840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6492411606609065,
+      "learning_rate": 7.46488495629913e-06,
+      "loss": 0.7453,
+      "step": 5841
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4196509139536047,
+      "learning_rate": 7.461698260591175e-06,
+      "loss": 0.668,
+      "step": 5842
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5833384209273522,
+      "learning_rate": 7.458511840376184e-06,
+      "loss": 0.767,
+      "step": 5843
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6025918147510856,
+      "learning_rate": 7.455325695999986e-06,
+      "loss": 0.7642,
+      "step": 5844
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6468470813377398,
+      "learning_rate": 7.452139827808389e-06,
+      "loss": 0.6987,
+      "step": 5845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8830976071957954,
+      "learning_rate": 7.44895423614717e-06,
+      "loss": 0.8303,
+      "step": 5846
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7994734743793335,
+      "learning_rate": 7.445768921362076e-06,
+      "loss": 0.7118,
+      "step": 5847
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7340069598124714,
+      "learning_rate": 7.442583883798822e-06,
+      "loss": 0.801,
+      "step": 5848
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7089345395030544,
+      "learning_rate": 7.439399123803091e-06,
+      "loss": 0.8203,
+      "step": 5849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4880693573774777,
+      "learning_rate": 7.436214641720545e-06,
+      "loss": 0.6633,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.559121371519373,
+      "learning_rate": 7.433030437896806e-06,
+      "loss": 0.6573,
+      "step": 5851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5023829504747404,
+      "learning_rate": 7.429846512677468e-06,
+      "loss": 0.7174,
+      "step": 5852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5674878085450052,
+      "learning_rate": 7.426662866408103e-06,
+      "loss": 0.6898,
+      "step": 5853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.619686042807621,
+      "learning_rate": 7.423479499434236e-06,
+      "loss": 0.656,
+      "step": 5854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.634263367846682,
+      "learning_rate": 7.42029641210138e-06,
+      "loss": 0.7641,
+      "step": 5855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6343792631487037,
+      "learning_rate": 7.4171136047550065e-06,
+      "loss": 0.6877,
+      "step": 5856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.42843883095044,
+      "learning_rate": 7.413931077740557e-06,
+      "loss": 0.604,
+      "step": 5857
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4921527424437566,
+      "learning_rate": 7.410748831403449e-06,
+      "loss": 0.6833,
+      "step": 5858
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6232802400330084,
+      "learning_rate": 7.4075668660890646e-06,
+      "loss": 0.744,
+      "step": 5859
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7416011582072546,
+      "learning_rate": 7.404385182142753e-06,
+      "loss": 0.7026,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6063603146265764,
+      "learning_rate": 7.4012037799098445e-06,
+      "loss": 0.6517,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8099053783680814,
+      "learning_rate": 7.39802265973562e-06,
+      "loss": 0.7787,
+      "step": 5862
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6263527004572946,
+      "learning_rate": 7.394841821965345e-06,
+      "loss": 0.7453,
+      "step": 5863
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4791392099068637,
+      "learning_rate": 7.391661266944254e-06,
+      "loss": 0.7141,
+      "step": 5864
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5802351106753338,
+      "learning_rate": 7.38848099501754e-06,
+      "loss": 0.7015,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4871552338594307,
+      "learning_rate": 7.385301006530378e-06,
+      "loss": 0.7324,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6085992046388986,
+      "learning_rate": 7.382121301827898e-06,
+      "loss": 0.743,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7153826304474897,
+      "learning_rate": 7.378941881255211e-06,
+      "loss": 0.7159,
+      "step": 5868
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8728107832178462,
+      "learning_rate": 7.3757627451573995e-06,
+      "loss": 0.5738,
+      "step": 5869
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5159368666910624,
+      "learning_rate": 7.372583893879499e-06,
+      "loss": 0.6209,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.449899387393714,
+      "learning_rate": 7.369405327766532e-06,
+      "loss": 0.6736,
+      "step": 5871
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6303463734462023,
+      "learning_rate": 7.366227047163476e-06,
+      "loss": 0.6618,
+      "step": 5872
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6863818917790458,
+      "learning_rate": 7.363049052415285e-06,
+      "loss": 0.836,
+      "step": 5873
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.820253340943577,
+      "learning_rate": 7.359871343866887e-06,
+      "loss": 0.741,
+      "step": 5874
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5931785045286797,
+      "learning_rate": 7.356693921863163e-06,
+      "loss": 0.6547,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6440388959898296,
+      "learning_rate": 7.3535167867489775e-06,
+      "loss": 0.7592,
+      "step": 5876
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5939645437591048,
+      "learning_rate": 7.350339938869162e-06,
+      "loss": 0.6622,
+      "step": 5877
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5597381779921151,
+      "learning_rate": 7.347163378568507e-06,
+      "loss": 0.8154,
+      "step": 5878
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4589897039416655,
+      "learning_rate": 7.343987106191786e-06,
+      "loss": 0.5976,
+      "step": 5879
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5984733750682367,
+      "learning_rate": 7.340811122083723e-06,
+      "loss": 0.6821,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4764646748866384,
+      "learning_rate": 7.3376354265890295e-06,
+      "loss": 0.7543,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5223731010955242,
+      "learning_rate": 7.334460020052379e-06,
+      "loss": 0.6846,
+      "step": 5882
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4660607880193885,
+      "learning_rate": 7.331284902818405e-06,
+      "loss": 0.6711,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.446061417819951,
+      "learning_rate": 7.328110075231725e-06,
+      "loss": 0.7085,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5524107193494978,
+      "learning_rate": 7.324935537636908e-06,
+      "loss": 0.697,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6555445285652144,
+      "learning_rate": 7.321761290378505e-06,
+      "loss": 0.7275,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4430592773775197,
+      "learning_rate": 7.318587333801036e-06,
+      "loss": 0.5604,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4908733198747264,
+      "learning_rate": 7.315413668248974e-06,
+      "loss": 0.7183,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5245766996522263,
+      "learning_rate": 7.312240294066782e-06,
+      "loss": 0.6907,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6246441521241848,
+      "learning_rate": 7.309067211598868e-06,
+      "loss": 0.7805,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.529185478581184,
+      "learning_rate": 7.305894421189628e-06,
+      "loss": 0.7304,
+      "step": 5891
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4448886607686946,
+      "learning_rate": 7.302721923183421e-06,
+      "loss": 0.6054,
+      "step": 5892
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5751107153435748,
+      "learning_rate": 7.299549717924565e-06,
+      "loss": 0.6843,
+      "step": 5893
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.718706567927009,
+      "learning_rate": 7.296377805757357e-06,
+      "loss": 0.6116,
+      "step": 5894
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5177722397037496,
+      "learning_rate": 7.29320618702606e-06,
+      "loss": 0.6166,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5444918903217328,
+      "learning_rate": 7.2900348620749016e-06,
+      "loss": 0.722,
+      "step": 5896
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.537921306975767,
+      "learning_rate": 7.286863831248078e-06,
+      "loss": 0.7269,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7285406302564792,
+      "learning_rate": 7.28369309488976e-06,
+      "loss": 0.7455,
+      "step": 5898
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5245798267270128,
+      "learning_rate": 7.280522653344076e-06,
+      "loss": 0.7289,
+      "step": 5899
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4914728486125917,
+      "learning_rate": 7.27735250695513e-06,
+      "loss": 0.6049,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.466868236257867,
+      "learning_rate": 7.274182656066992e-06,
+      "loss": 0.6538,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7183062480640914,
+      "learning_rate": 7.271013101023702e-06,
+      "loss": 0.7655,
+      "step": 5902
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8887092239674945,
+      "learning_rate": 7.26784384216926e-06,
+      "loss": 0.8537,
+      "step": 5903
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7400780102568585,
+      "learning_rate": 7.264674879847644e-06,
+      "loss": 0.7871,
+      "step": 5904
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7134581127166453,
+      "learning_rate": 7.261506214402796e-06,
+      "loss": 0.7119,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5867941278621458,
+      "learning_rate": 7.258337846178621e-06,
+      "loss": 0.6933,
+      "step": 5906
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5424375073206111,
+      "learning_rate": 7.255169775519e-06,
+      "loss": 0.8146,
+      "step": 5907
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3814753377772704,
+      "learning_rate": 7.2520020027677715e-06,
+      "loss": 0.6826,
+      "step": 5908
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6083435545153657,
+      "learning_rate": 7.248834528268756e-06,
+      "loss": 0.7188,
+      "step": 5909
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6591768526517174,
+      "learning_rate": 7.245667352365727e-06,
+      "loss": 0.7414,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4171103717364297,
+      "learning_rate": 7.242500475402433e-06,
+      "loss": 0.6912,
+      "step": 5911
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5330338232654563,
+      "learning_rate": 7.239333897722591e-06,
+      "loss": 0.7653,
+      "step": 5912
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5562892061873899,
+      "learning_rate": 7.2361676196698834e-06,
+      "loss": 0.6745,
+      "step": 5913
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6240138314562722,
+      "learning_rate": 7.233001641587958e-06,
+      "loss": 0.5879,
+      "step": 5914
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.617474529996535,
+      "learning_rate": 7.229835963820435e-06,
+      "loss": 0.8026,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6839525778259303,
+      "learning_rate": 7.226670586710896e-06,
+      "loss": 0.7293,
+      "step": 5916
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5312147309286948,
+      "learning_rate": 7.223505510602893e-06,
+      "loss": 0.6628,
+      "step": 5917
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7960645410763412,
+      "learning_rate": 7.220340735839953e-06,
+      "loss": 0.6784,
+      "step": 5918
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.700530965726789,
+      "learning_rate": 7.217176262765551e-06,
+      "loss": 0.746,
+      "step": 5919
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.77136543867775,
+      "learning_rate": 7.2140120917231525e-06,
+      "loss": 0.728,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6255467952292564,
+      "learning_rate": 7.210848223056169e-06,
+      "loss": 0.6546,
+      "step": 5921
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5471405337399904,
+      "learning_rate": 7.207684657107994e-06,
+      "loss": 0.7178,
+      "step": 5922
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4928474424652634,
+      "learning_rate": 7.204521394221986e-06,
+      "loss": 0.7027,
+      "step": 5923
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5710364609704472,
+      "learning_rate": 7.201358434741461e-06,
+      "loss": 0.7739,
+      "step": 5924
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5515633009029757,
+      "learning_rate": 7.1981957790097155e-06,
+      "loss": 0.6327,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.589661024787032,
+      "learning_rate": 7.195033427369998e-06,
+      "loss": 0.8063,
+      "step": 5926
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7548994446950992,
+      "learning_rate": 7.191871380165538e-06,
+      "loss": 0.7202,
+      "step": 5927
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5781662873618456,
+      "learning_rate": 7.1887096377395305e-06,
+      "loss": 0.6457,
+      "step": 5928
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6275642291493222,
+      "learning_rate": 7.185548200435123e-06,
+      "loss": 0.7623,
+      "step": 5929
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8226216363463947,
+      "learning_rate": 7.182387068595445e-06,
+      "loss": 0.6989,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6337489744342286,
+      "learning_rate": 7.179226242563593e-06,
+      "loss": 0.8565,
+      "step": 5931
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7040405735331745,
+      "learning_rate": 7.176065722682616e-06,
+      "loss": 0.6508,
+      "step": 5932
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7053161007144007,
+      "learning_rate": 7.172905509295547e-06,
+      "loss": 0.6041,
+      "step": 5933
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5160666240079106,
+      "learning_rate": 7.1697456027453705e-06,
+      "loss": 0.7378,
+      "step": 5934
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5134763013492174,
+      "learning_rate": 7.166586003375049e-06,
+      "loss": 0.6613,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5604089079300418,
+      "learning_rate": 7.16342671152751e-06,
+      "loss": 0.7043,
+      "step": 5936
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.705865897955065,
+      "learning_rate": 7.16026772754564e-06,
+      "loss": 0.7221,
+      "step": 5937
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5183849133745817,
+      "learning_rate": 7.157109051772304e-06,
+      "loss": 0.6544,
+      "step": 5938
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6388424865238325,
+      "learning_rate": 7.153950684550317e-06,
+      "loss": 0.6588,
+      "step": 5939
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.735175336027495,
+      "learning_rate": 7.150792626222476e-06,
+      "loss": 0.8374,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4780545701966636,
+      "learning_rate": 7.147634877131544e-06,
+      "loss": 0.7295,
+      "step": 5941
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.635128129316677,
+      "learning_rate": 7.144477437620235e-06,
+      "loss": 0.7532,
+      "step": 5942
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.531452447180397,
+      "learning_rate": 7.14132030803125e-06,
+      "loss": 0.7193,
+      "step": 5943
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6133299287122889,
+      "learning_rate": 7.138163488707235e-06,
+      "loss": 0.7288,
+      "step": 5944
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.615628410179108,
+      "learning_rate": 7.1350069799908205e-06,
+      "loss": 0.6419,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4970890926910432,
+      "learning_rate": 7.131850782224598e-06,
+      "loss": 0.7031,
+      "step": 5946
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.479217347173918,
+      "learning_rate": 7.128694895751118e-06,
+      "loss": 0.709,
+      "step": 5947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4703328917956562,
+      "learning_rate": 7.1255393209129034e-06,
+      "loss": 0.6481,
+      "step": 5948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5584138647975212,
+      "learning_rate": 7.1223840580524485e-06,
+      "loss": 0.6695,
+      "step": 5949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5948021840095914,
+      "learning_rate": 7.1192291075121985e-06,
+      "loss": 0.7755,
+      "step": 5950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.589248305768824,
+      "learning_rate": 7.116074469634582e-06,
+      "loss": 0.8458,
+      "step": 5951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.685088414455333,
+      "learning_rate": 7.112920144761981e-06,
+      "loss": 0.7741,
+      "step": 5952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6519851958314915,
+      "learning_rate": 7.109766133236747e-06,
+      "loss": 0.7756,
+      "step": 5953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.603902906380497,
+      "learning_rate": 7.106612435401204e-06,
+      "loss": 0.7936,
+      "step": 5954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5540901955171227,
+      "learning_rate": 7.103459051597634e-06,
+      "loss": 0.6245,
+      "step": 5955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4938465866104673,
+      "learning_rate": 7.100305982168287e-06,
+      "loss": 0.7651,
+      "step": 5956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4482093578275226,
+      "learning_rate": 7.097153227455379e-06,
+      "loss": 0.7077,
+      "step": 5957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.458455597539955,
+      "learning_rate": 7.09400078780109e-06,
+      "loss": 0.6886,
+      "step": 5958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6426093964650683,
+      "learning_rate": 7.090848663547574e-06,
+      "loss": 0.7427,
+      "step": 5959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5489288651190687,
+      "learning_rate": 7.087696855036941e-06,
+      "loss": 0.7297,
+      "step": 5960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4600247920573843,
+      "learning_rate": 7.084545362611271e-06,
+      "loss": 0.6347,
+      "step": 5961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.45701917668946,
+      "learning_rate": 7.081394186612607e-06,
+      "loss": 0.7495,
+      "step": 5962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7218090265097128,
+      "learning_rate": 7.078243327382965e-06,
+      "loss": 0.7223,
+      "step": 5963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.667199490568043,
+      "learning_rate": 7.075092785264318e-06,
+      "loss": 0.8159,
+      "step": 5964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6435345957610008,
+      "learning_rate": 7.071942560598607e-06,
+      "loss": 0.6628,
+      "step": 5965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.512973675003832,
+      "learning_rate": 7.068792653727745e-06,
+      "loss": 0.6767,
+      "step": 5966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.715175169843844,
+      "learning_rate": 7.065643064993598e-06,
+      "loss": 0.7698,
+      "step": 5967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.695624356046342,
+      "learning_rate": 7.062493794738008e-06,
+      "loss": 0.7545,
+      "step": 5968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6865151048997264,
+      "learning_rate": 7.059344843302783e-06,
+      "loss": 0.7827,
+      "step": 5969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7368175081011257,
+      "learning_rate": 7.0561962110296845e-06,
+      "loss": 0.7576,
+      "step": 5970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.615677836368483,
+      "learning_rate": 7.0530478982604524e-06,
+      "loss": 0.657,
+      "step": 5971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.693141974058053,
+      "learning_rate": 7.049899905336788e-06,
+      "loss": 0.7445,
+      "step": 5972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5262126051228404,
+      "learning_rate": 7.046752232600351e-06,
+      "loss": 0.6724,
+      "step": 5973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7065540618543513,
+      "learning_rate": 7.043604880392781e-06,
+      "loss": 0.7284,
+      "step": 5974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5338232721417555,
+      "learning_rate": 7.0404578490556616e-06,
+      "loss": 0.6668,
+      "step": 5975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5935989717722543,
+      "learning_rate": 7.037311138930563e-06,
+      "loss": 0.6522,
+      "step": 5976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6575930100117833,
+      "learning_rate": 7.034164750359011e-06,
+      "loss": 0.627,
+      "step": 5977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7356854343165955,
+      "learning_rate": 7.031018683682494e-06,
+      "loss": 0.7598,
+      "step": 5978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6779200410679829,
+      "learning_rate": 7.02787293924247e-06,
+      "loss": 0.7907,
+      "step": 5979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.501651087009643,
+      "learning_rate": 7.024727517380356e-06,
+      "loss": 0.6912,
+      "step": 5980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5649323303119,
+      "learning_rate": 7.021582418437543e-06,
+      "loss": 0.6899,
+      "step": 5981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6378941911076537,
+      "learning_rate": 7.018437642755384e-06,
+      "loss": 0.6915,
+      "step": 5982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5636475281530569,
+      "learning_rate": 7.015293190675191e-06,
+      "loss": 0.6637,
+      "step": 5983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.561748989419727,
+      "learning_rate": 7.012149062538249e-06,
+      "loss": 0.7508,
+      "step": 5984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.75746553351738,
+      "learning_rate": 7.009005258685797e-06,
+      "loss": 0.6662,
+      "step": 5985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7962183540031025,
+      "learning_rate": 7.005861779459051e-06,
+      "loss": 0.674,
+      "step": 5986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5312431132527884,
+      "learning_rate": 7.00271862519919e-06,
+      "loss": 0.6533,
+      "step": 5987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8068991522290494,
+      "learning_rate": 6.999575796247344e-06,
+      "loss": 0.7689,
+      "step": 5988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4364228321091383,
+      "learning_rate": 6.996433292944627e-06,
+      "loss": 0.759,
+      "step": 5989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5059840142036114,
+      "learning_rate": 6.993291115632108e-06,
+      "loss": 0.7712,
+      "step": 5990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.879677897781722,
+      "learning_rate": 6.990149264650814e-06,
+      "loss": 0.735,
+      "step": 5991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5384761249987013,
+      "learning_rate": 6.987007740341754e-06,
+      "loss": 0.6578,
+      "step": 5992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.442254383864038,
+      "learning_rate": 6.983866543045881e-06,
+      "loss": 0.6452,
+      "step": 5993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5261177107975168,
+      "learning_rate": 6.980725673104128e-06,
+      "loss": 0.708,
+      "step": 5994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5038665795578166,
+      "learning_rate": 6.97758513085739e-06,
+      "loss": 0.7081,
+      "step": 5995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.846541150792894,
+      "learning_rate": 6.974444916646517e-06,
+      "loss": 0.7867,
+      "step": 5996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.690342301741687,
+      "learning_rate": 6.971305030812339e-06,
+      "loss": 0.7515,
+      "step": 5997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5635055628794958,
+      "learning_rate": 6.968165473695632e-06,
+      "loss": 0.6293,
+      "step": 5998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4455948190504697,
+      "learning_rate": 6.96502624563715e-06,
+      "loss": 0.6306,
+      "step": 5999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.585497684641971,
+      "learning_rate": 6.961887346977611e-06,
+      "loss": 0.6823,
+      "step": 6000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4864011580113403,
+      "learning_rate": 6.958748778057687e-06,
+      "loss": 0.7411,
+      "step": 6001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7449596125569689,
+      "learning_rate": 6.955610539218023e-06,
+      "loss": 0.6064,
+      "step": 6002
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6330422724486302,
+      "learning_rate": 6.952472630799227e-06,
+      "loss": 0.7924,
+      "step": 6003
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.561676435086924,
+      "learning_rate": 6.949335053141868e-06,
+      "loss": 0.6949,
+      "step": 6004
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5681058178598866,
+      "learning_rate": 6.9461978065864835e-06,
+      "loss": 0.8036,
+      "step": 6005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5518107210158107,
+      "learning_rate": 6.943060891473572e-06,
+      "loss": 0.7636,
+      "step": 6006
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7040411881612014,
+      "learning_rate": 6.939924308143591e-06,
+      "loss": 0.7985,
+      "step": 6007
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.673280397750217,
+      "learning_rate": 6.936788056936976e-06,
+      "loss": 0.7416,
+      "step": 6008
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.674118049602933,
+      "learning_rate": 6.933652138194114e-06,
+      "loss": 0.6395,
+      "step": 6009
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5361770186226622,
+      "learning_rate": 6.93051655225536e-06,
+      "loss": 0.6385,
+      "step": 6010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5032438867630116,
+      "learning_rate": 6.9273812994610315e-06,
+      "loss": 0.6813,
+      "step": 6011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7101399111295492,
+      "learning_rate": 6.924246380151411e-06,
+      "loss": 0.8143,
+      "step": 6012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5441825955179973,
+      "learning_rate": 6.92111179466675e-06,
+      "loss": 0.6683,
+      "step": 6013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7242217401584956,
+      "learning_rate": 6.917977543347254e-06,
+      "loss": 0.7674,
+      "step": 6014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7117024400753804,
+      "learning_rate": 6.914843626533099e-06,
+      "loss": 0.7421,
+      "step": 6015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.644187079438976,
+      "learning_rate": 6.911710044564419e-06,
+      "loss": 0.769,
+      "step": 6016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4185821366042153,
+      "learning_rate": 6.908576797781321e-06,
+      "loss": 0.7162,
+      "step": 6017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5664785236770578,
+      "learning_rate": 6.905443886523868e-06,
+      "loss": 0.6154,
+      "step": 6018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.472890070195125,
+      "learning_rate": 6.902311311132084e-06,
+      "loss": 0.6228,
+      "step": 6019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6096312989951134,
+      "learning_rate": 6.89917907194597e-06,
+      "loss": 0.6971,
+      "step": 6020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7037258038186427,
+      "learning_rate": 6.896047169305471e-06,
+      "loss": 0.661,
+      "step": 6021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5649396853725739,
+      "learning_rate": 6.892915603550512e-06,
+      "loss": 0.634,
+      "step": 6022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6722121537895382,
+      "learning_rate": 6.8897843750209796e-06,
+      "loss": 0.7228,
+      "step": 6023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5956272085005472,
+      "learning_rate": 6.8866534840567104e-06,
+      "loss": 0.7118,
+      "step": 6024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.599425073147109,
+      "learning_rate": 6.883522930997517e-06,
+      "loss": 0.7629,
+      "step": 6025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6284369870991078,
+      "learning_rate": 6.88039271618318e-06,
+      "loss": 0.7095,
+      "step": 6026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4076269843619167,
+      "learning_rate": 6.877262839953422e-06,
+      "loss": 0.5916,
+      "step": 6027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5691541798606565,
+      "learning_rate": 6.874133302647953e-06,
+      "loss": 0.6988,
+      "step": 6028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5347110368511394,
+      "learning_rate": 6.871004104606427e-06,
+      "loss": 0.6454,
+      "step": 6029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4321167031271145,
+      "learning_rate": 6.867875246168474e-06,
+      "loss": 0.6535,
+      "step": 6030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6411822049037168,
+      "learning_rate": 6.864746727673685e-06,
+      "loss": 0.6945,
+      "step": 6031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.633086470213813,
+      "learning_rate": 6.861618549461606e-06,
+      "loss": 0.624,
+      "step": 6032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6858246877293717,
+      "learning_rate": 6.858490711871759e-06,
+      "loss": 0.686,
+      "step": 6033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5239869152303707,
+      "learning_rate": 6.855363215243612e-06,
+      "loss": 0.8371,
+      "step": 6034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.445563473951057,
+      "learning_rate": 6.852236059916612e-06,
+      "loss": 0.6643,
+      "step": 6035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6714240174258932,
+      "learning_rate": 6.849109246230167e-06,
+      "loss": 0.7424,
+      "step": 6036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.683585445655326,
+      "learning_rate": 6.845982774523634e-06,
+      "loss": 0.6813,
+      "step": 6037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6850963323083905,
+      "learning_rate": 6.842856645136351e-06,
+      "loss": 0.744,
+      "step": 6038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6178748682318473,
+      "learning_rate": 6.839730858407604e-06,
+      "loss": 0.6569,
+      "step": 6039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5032196567724914,
+      "learning_rate": 6.836605414676652e-06,
+      "loss": 0.6416,
+      "step": 6040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5419909852187603,
+      "learning_rate": 6.8334803142827144e-06,
+      "loss": 0.7778,
+      "step": 6041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6887785419890449,
+      "learning_rate": 6.830355557564966e-06,
+      "loss": 0.8648,
+      "step": 6042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4451232947180794,
+      "learning_rate": 6.827231144862555e-06,
+      "loss": 0.7358,
+      "step": 6043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.660669903928941,
+      "learning_rate": 6.82410707651459e-06,
+      "loss": 0.7228,
+      "step": 6044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6440758805537505,
+      "learning_rate": 6.820983352860133e-06,
+      "loss": 0.7577,
+      "step": 6045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6257757580594896,
+      "learning_rate": 6.817859974238223e-06,
+      "loss": 0.6243,
+      "step": 6046
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6584220789207265,
+      "learning_rate": 6.814736940987845e-06,
+      "loss": 0.6615,
+      "step": 6047
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5600593702346273,
+      "learning_rate": 6.81161425344796e-06,
+      "loss": 0.6646,
+      "step": 6048
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7382755089642217,
+      "learning_rate": 6.808491911957492e-06,
+      "loss": 0.8191,
+      "step": 6049
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6370998175010671,
+      "learning_rate": 6.805369916855313e-06,
+      "loss": 0.8585,
+      "step": 6050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5579772194685,
+      "learning_rate": 6.8022482684802745e-06,
+      "loss": 0.6841,
+      "step": 6051
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5781599509351145,
+      "learning_rate": 6.799126967171177e-06,
+      "loss": 0.7058,
+      "step": 6052
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6306166871228853,
+      "learning_rate": 6.7960060132667895e-06,
+      "loss": 0.7381,
+      "step": 6053
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8356173142033885,
+      "learning_rate": 6.792885407105848e-06,
+      "loss": 0.7518,
+      "step": 6054
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6522651945471996,
+      "learning_rate": 6.789765149027039e-06,
+      "loss": 0.6215,
+      "step": 6055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.53831203818576,
+      "learning_rate": 6.786645239369022e-06,
+      "loss": 0.7443,
+      "step": 6056
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9292567739728215,
+      "learning_rate": 6.783525678470413e-06,
+      "loss": 0.7387,
+      "step": 6057
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3837803608137365,
+      "learning_rate": 6.78040646666979e-06,
+      "loss": 0.7023,
+      "step": 6058
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6306535740518562,
+      "learning_rate": 6.777287604305698e-06,
+      "loss": 0.7212,
+      "step": 6059
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6551843866151499,
+      "learning_rate": 6.774169091716638e-06,
+      "loss": 0.796,
+      "step": 6060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7677581368413209,
+      "learning_rate": 6.771050929241076e-06,
+      "loss": 0.8101,
+      "step": 6061
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5152515724530302,
+      "learning_rate": 6.767933117217444e-06,
+      "loss": 0.7826,
+      "step": 6062
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7830840121583593,
+      "learning_rate": 6.764815655984125e-06,
+      "loss": 0.7526,
+      "step": 6063
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7562929056380636,
+      "learning_rate": 6.761698545879476e-06,
+      "loss": 0.7104,
+      "step": 6064
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5567240302854402,
+      "learning_rate": 6.758581787241807e-06,
+      "loss": 0.5608,
+      "step": 6065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.723317681957902,
+      "learning_rate": 6.755465380409394e-06,
+      "loss": 0.688,
+      "step": 6066
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5952815129509346,
+      "learning_rate": 6.7523493257204776e-06,
+      "loss": 0.7464,
+      "step": 6067
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4090609686958033,
+      "learning_rate": 6.749233623513254e-06,
+      "loss": 0.7506,
+      "step": 6068
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5473101411992871,
+      "learning_rate": 6.746118274125883e-06,
+      "loss": 0.6807,
+      "step": 6069
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6216043388355958,
+      "learning_rate": 6.743003277896487e-06,
+      "loss": 0.6917,
+      "step": 6070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.916644081908343,
+      "learning_rate": 6.739888635163155e-06,
+      "loss": 0.7651,
+      "step": 6071
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5161974751902205,
+      "learning_rate": 6.7367743462639265e-06,
+      "loss": 0.6257,
+      "step": 6072
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.300149205632678,
+      "learning_rate": 6.733660411536811e-06,
+      "loss": 0.6334,
+      "step": 6073
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.553990093872253,
+      "learning_rate": 6.7305468313197815e-06,
+      "loss": 0.7501,
+      "step": 6074
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5132042880834147,
+      "learning_rate": 6.7274336059507614e-06,
+      "loss": 0.7175,
+      "step": 6075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5129442394667727,
+      "learning_rate": 6.724320735767646e-06,
+      "loss": 0.708,
+      "step": 6076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.51569068130344,
+      "learning_rate": 6.721208221108293e-06,
+      "loss": 0.7555,
+      "step": 6077
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6666093513562708,
+      "learning_rate": 6.71809606231051e-06,
+      "loss": 0.7786,
+      "step": 6078
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3740089472688364,
+      "learning_rate": 6.714984259712074e-06,
+      "loss": 0.6608,
+      "step": 6079
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6533318917815236,
+      "learning_rate": 6.71187281365073e-06,
+      "loss": 0.7125,
+      "step": 6080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5660729332736263,
+      "learning_rate": 6.708761724464168e-06,
+      "loss": 0.6867,
+      "step": 6081
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5812173362275896,
+      "learning_rate": 6.705650992490054e-06,
+      "loss": 0.6205,
+      "step": 6082
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4441682341203965,
+      "learning_rate": 6.7025406180660046e-06,
+      "loss": 0.6858,
+      "step": 6083
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4743755511325085,
+      "learning_rate": 6.699430601529604e-06,
+      "loss": 0.6628,
+      "step": 6084
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6450244119277522,
+      "learning_rate": 6.696320943218401e-06,
+      "loss": 0.6891,
+      "step": 6085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5033342990472907,
+      "learning_rate": 6.69321164346989e-06,
+      "loss": 0.7987,
+      "step": 6086
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6742692228987246,
+      "learning_rate": 6.690102702621548e-06,
+      "loss": 0.7311,
+      "step": 6087
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6586419641837007,
+      "learning_rate": 6.686994121010794e-06,
+      "loss": 0.6917,
+      "step": 6088
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5909589091117324,
+      "learning_rate": 6.683885898975016e-06,
+      "loss": 0.6332,
+      "step": 6089
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.450643132802545,
+      "learning_rate": 6.680778036851572e-06,
+      "loss": 0.6433,
+      "step": 6090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.551984481934261,
+      "learning_rate": 6.677670534977759e-06,
+      "loss": 0.7499,
+      "step": 6091
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5859172831013613,
+      "learning_rate": 6.674563393690858e-06,
+      "loss": 0.7056,
+      "step": 6092
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6366553804795465,
+      "learning_rate": 6.6714566133280944e-06,
+      "loss": 0.6751,
+      "step": 6093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6655711827187523,
+      "learning_rate": 6.668350194226662e-06,
+      "loss": 0.7645,
+      "step": 6094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5485819950508424,
+      "learning_rate": 6.665244136723719e-06,
+      "loss": 0.727,
+      "step": 6095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4127253810983589,
+      "learning_rate": 6.662138441156371e-06,
+      "loss": 0.6536,
+      "step": 6096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6329594626580382,
+      "learning_rate": 6.659033107861697e-06,
+      "loss": 0.7044,
+      "step": 6097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7760481582902228,
+      "learning_rate": 6.655928137176735e-06,
+      "loss": 0.7399,
+      "step": 6098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6032009667240732,
+      "learning_rate": 6.652823529438476e-06,
+      "loss": 0.6502,
+      "step": 6099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4651867018769336,
+      "learning_rate": 6.649719284983882e-06,
+      "loss": 0.7245,
+      "step": 6100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5920615165314944,
+      "learning_rate": 6.646615404149863e-06,
+      "loss": 0.7663,
+      "step": 6101
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6675682691281073,
+      "learning_rate": 6.6435118872733016e-06,
+      "loss": 0.7442,
+      "step": 6102
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6036547915026613,
+      "learning_rate": 6.640408734691039e-06,
+      "loss": 0.7039,
+      "step": 6103
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.702638743603976,
+      "learning_rate": 6.637305946739865e-06,
+      "loss": 0.7123,
+      "step": 6104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6000778248442582,
+      "learning_rate": 6.63420352375655e-06,
+      "loss": 0.6588,
+      "step": 6105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5714725149545097,
+      "learning_rate": 6.631101466077801e-06,
+      "loss": 0.677,
+      "step": 6106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6820828931724316,
+      "learning_rate": 6.627999774040305e-06,
+      "loss": 0.7083,
+      "step": 6107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4645956316659832,
+      "learning_rate": 6.624898447980706e-06,
+      "loss": 0.6704,
+      "step": 6108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6712984172262535,
+      "learning_rate": 6.6217974882355955e-06,
+      "loss": 0.6464,
+      "step": 6109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.570087609165966,
+      "learning_rate": 6.618696895141541e-06,
+      "loss": 0.7499,
+      "step": 6110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7519649038605543,
+      "learning_rate": 6.615596669035059e-06,
+      "loss": 0.6598,
+      "step": 6111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5637128365097759,
+      "learning_rate": 6.6124968102526325e-06,
+      "loss": 0.8167,
+      "step": 6112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8123296526525825,
+      "learning_rate": 6.6093973191307055e-06,
+      "loss": 0.6705,
+      "step": 6113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5613541223195324,
+      "learning_rate": 6.606298196005673e-06,
+      "loss": 0.7487,
+      "step": 6114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7326002131841742,
+      "learning_rate": 6.603199441213901e-06,
+      "loss": 0.7714,
+      "step": 6115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7424750174350074,
+      "learning_rate": 6.60010105509171e-06,
+      "loss": 0.7891,
+      "step": 6116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5365332328220989,
+      "learning_rate": 6.597003037975379e-06,
+      "loss": 0.7173,
+      "step": 6117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5147365728214297,
+      "learning_rate": 6.593905390201154e-06,
+      "loss": 0.6687,
+      "step": 6118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6877996988645476,
+      "learning_rate": 6.590808112105232e-06,
+      "loss": 0.8397,
+      "step": 6119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7092597152986675,
+      "learning_rate": 6.5877112040237746e-06,
+      "loss": 0.7818,
+      "step": 6120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.613334053324412,
+      "learning_rate": 6.584614666292906e-06,
+      "loss": 0.717,
+      "step": 6121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5180147856705175,
+      "learning_rate": 6.581518499248705e-06,
+      "loss": 0.7353,
+      "step": 6122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5256391495854664,
+      "learning_rate": 6.578422703227211e-06,
+      "loss": 0.7384,
+      "step": 6123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6215104435427263,
+      "learning_rate": 6.5753272785644225e-06,
+      "loss": 0.5765,
+      "step": 6124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4798270566748801,
+      "learning_rate": 6.572232225596306e-06,
+      "loss": 0.636,
+      "step": 6125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6553762148182485,
+      "learning_rate": 6.569137544658775e-06,
+      "loss": 0.8014,
+      "step": 6126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.621778896682407,
+      "learning_rate": 6.56604323608771e-06,
+      "loss": 0.6951,
+      "step": 6127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5478542582256125,
+      "learning_rate": 6.562949300218955e-06,
+      "loss": 0.6891,
+      "step": 6128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.751566216790655,
+      "learning_rate": 6.559855737388299e-06,
+      "loss": 0.7081,
+      "step": 6129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6892966354547168,
+      "learning_rate": 6.556762547931505e-06,
+      "loss": 0.7547,
+      "step": 6130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4676227029041156,
+      "learning_rate": 6.553669732184296e-06,
+      "loss": 0.6531,
+      "step": 6131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5368957217679278,
+      "learning_rate": 6.550577290482336e-06,
+      "loss": 0.6174,
+      "step": 6132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8966492736331908,
+      "learning_rate": 6.547485223161274e-06,
+      "loss": 0.8064,
+      "step": 6133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6919571246987175,
+      "learning_rate": 6.544393530556697e-06,
+      "loss": 0.7939,
+      "step": 6134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5433280990541385,
+      "learning_rate": 6.54130221300416e-06,
+      "loss": 0.6668,
+      "step": 6135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4384780138474969,
+      "learning_rate": 6.538211270839185e-06,
+      "loss": 0.6493,
+      "step": 6136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.483118398389075,
+      "learning_rate": 6.535120704397236e-06,
+      "loss": 0.6313,
+      "step": 6137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5126790622807804,
+      "learning_rate": 6.53203051401375e-06,
+      "loss": 0.6973,
+      "step": 6138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7201848995408757,
+      "learning_rate": 6.528940700024122e-06,
+      "loss": 0.7339,
+      "step": 6139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7302480571234962,
+      "learning_rate": 6.525851262763696e-06,
+      "loss": 0.666,
+      "step": 6140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7036931968958287,
+      "learning_rate": 6.522762202567792e-06,
+      "loss": 0.7611,
+      "step": 6141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.609127038666171,
+      "learning_rate": 6.519673519771666e-06,
+      "loss": 0.6689,
+      "step": 6142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4781757504167778,
+      "learning_rate": 6.516585214710554e-06,
+      "loss": 0.6643,
+      "step": 6143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5196067810040579,
+      "learning_rate": 6.513497287719648e-06,
+      "loss": 0.5777,
+      "step": 6144
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5785347540072658,
+      "learning_rate": 6.510409739134082e-06,
+      "loss": 0.7492,
+      "step": 6145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.666016850977178,
+      "learning_rate": 6.507322569288974e-06,
+      "loss": 0.7188,
+      "step": 6146
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6450134972059112,
+      "learning_rate": 6.504235778519376e-06,
+      "loss": 0.6529,
+      "step": 6147
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.502261606984428,
+      "learning_rate": 6.501149367160319e-06,
+      "loss": 0.6806,
+      "step": 6148
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5076422897426254,
+      "learning_rate": 6.4980633355467845e-06,
+      "loss": 0.7812,
+      "step": 6149
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4866092473823802,
+      "learning_rate": 6.494977684013708e-06,
+      "loss": 0.5794,
+      "step": 6150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5008093258540893,
+      "learning_rate": 6.4918924128959945e-06,
+      "loss": 0.6753,
+      "step": 6151
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7265260554302626,
+      "learning_rate": 6.488807522528496e-06,
+      "loss": 0.7154,
+      "step": 6152
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5887570645077855,
+      "learning_rate": 6.48572301324603e-06,
+      "loss": 0.7191,
+      "step": 6153
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5810627427814312,
+      "learning_rate": 6.482638885383379e-06,
+      "loss": 0.6966,
+      "step": 6154
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6789536547200257,
+      "learning_rate": 6.479555139275267e-06,
+      "loss": 0.7727,
+      "step": 6155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6242882408342163,
+      "learning_rate": 6.476471775256391e-06,
+      "loss": 0.6108,
+      "step": 6156
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6112020057308443,
+      "learning_rate": 6.473388793661406e-06,
+      "loss": 0.6523,
+      "step": 6157
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0051813216712024,
+      "learning_rate": 6.470306194824914e-06,
+      "loss": 0.7441,
+      "step": 6158
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6806981253531708,
+      "learning_rate": 6.467223979081491e-06,
+      "loss": 0.7096,
+      "step": 6159
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.702782121668881,
+      "learning_rate": 6.464142146765653e-06,
+      "loss": 0.7192,
+      "step": 6160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6169778774789403,
+      "learning_rate": 6.4610606982118914e-06,
+      "loss": 0.7052,
+      "step": 6161
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6807203061396876,
+      "learning_rate": 6.457979633754652e-06,
+      "loss": 0.7438,
+      "step": 6162
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6349504107469093,
+      "learning_rate": 6.454898953728328e-06,
+      "loss": 0.7091,
+      "step": 6163
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.562430195567715,
+      "learning_rate": 6.4518186584672856e-06,
+      "loss": 0.7275,
+      "step": 6164
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6464753506977141,
+      "learning_rate": 6.448738748305841e-06,
+      "loss": 0.6985,
+      "step": 6165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6695608842141376,
+      "learning_rate": 6.445659223578267e-06,
+      "loss": 0.6286,
+      "step": 6166
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5903742512985368,
+      "learning_rate": 6.442580084618806e-06,
+      "loss": 0.7277,
+      "step": 6167
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6137454812957188,
+      "learning_rate": 6.439501331761639e-06,
+      "loss": 0.6366,
+      "step": 6168
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4440989137214413,
+      "learning_rate": 6.436422965340925e-06,
+      "loss": 0.6553,
+      "step": 6169
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6574414923263707,
+      "learning_rate": 6.4333449856907705e-06,
+      "loss": 0.7672,
+      "step": 6170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.458801356848339,
+      "learning_rate": 6.43026739314524e-06,
+      "loss": 0.6427,
+      "step": 6171
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5653358507579374,
+      "learning_rate": 6.427190188038362e-06,
+      "loss": 0.7527,
+      "step": 6172
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.650511862021473,
+      "learning_rate": 6.424113370704115e-06,
+      "loss": 0.7734,
+      "step": 6173
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5755765541554974,
+      "learning_rate": 6.421036941476439e-06,
+      "loss": 0.6524,
+      "step": 6174
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5420151974503422,
+      "learning_rate": 6.417960900689238e-06,
+      "loss": 0.6917,
+      "step": 6175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7208568860923064,
+      "learning_rate": 6.414885248676361e-06,
+      "loss": 0.7065,
+      "step": 6176
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6013565217810959,
+      "learning_rate": 6.411809985771629e-06,
+      "loss": 0.73,
+      "step": 6177
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5697716167814815,
+      "learning_rate": 6.408735112308806e-06,
+      "loss": 0.6028,
+      "step": 6178
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.571569770741846,
+      "learning_rate": 6.405660628621628e-06,
+      "loss": 0.7557,
+      "step": 6179
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.40999018231601,
+      "learning_rate": 6.40258653504378e-06,
+      "loss": 0.6783,
+      "step": 6180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3555682140849614,
+      "learning_rate": 6.3995128319089036e-06,
+      "loss": 0.6635,
+      "step": 6181
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7179220579329704,
+      "learning_rate": 6.396439519550609e-06,
+      "loss": 0.821,
+      "step": 6182
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.556213439686506,
+      "learning_rate": 6.3933665983024465e-06,
+      "loss": 0.7582,
+      "step": 6183
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.76363791750226,
+      "learning_rate": 6.390294068497937e-06,
+      "loss": 0.7535,
+      "step": 6184
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7217112711922193,
+      "learning_rate": 6.387221930470564e-06,
+      "loss": 0.6494,
+      "step": 6185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.639820890694774,
+      "learning_rate": 6.3841501845537464e-06,
+      "loss": 0.7747,
+      "step": 6186
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8380696332977846,
+      "learning_rate": 6.3810788310808855e-06,
+      "loss": 0.7267,
+      "step": 6187
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6150044913304795,
+      "learning_rate": 6.378007870385321e-06,
+      "loss": 0.7976,
+      "step": 6188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6080626517186347,
+      "learning_rate": 6.3749373028003595e-06,
+      "loss": 0.645,
+      "step": 6189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.590231884761789,
+      "learning_rate": 6.371867128659267e-06,
+      "loss": 0.6826,
+      "step": 6190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6438217494955731,
+      "learning_rate": 6.368797348295257e-06,
+      "loss": 0.7977,
+      "step": 6191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7206786971877635,
+      "learning_rate": 6.36572796204151e-06,
+      "loss": 0.7241,
+      "step": 6192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6000650003086683,
+      "learning_rate": 6.362658970231161e-06,
+      "loss": 0.7469,
+      "step": 6193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.543427534178332,
+      "learning_rate": 6.3595903731972975e-06,
+      "loss": 0.7506,
+      "step": 6194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6441500222599714,
+      "learning_rate": 6.356522171272972e-06,
+      "loss": 0.7373,
+      "step": 6195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5579041192143457,
+      "learning_rate": 6.353454364791184e-06,
+      "loss": 0.7233,
+      "step": 6196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6165157378519235,
+      "learning_rate": 6.350386954084898e-06,
+      "loss": 0.7153,
+      "step": 6197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6553996838473866,
+      "learning_rate": 6.34731993948704e-06,
+      "loss": 0.6837,
+      "step": 6198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6509896130097286,
+      "learning_rate": 6.344253321330476e-06,
+      "loss": 0.7204,
+      "step": 6199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5788859467893934,
+      "learning_rate": 6.341187099948049e-06,
+      "loss": 0.7383,
+      "step": 6200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.740660603581767,
+      "learning_rate": 6.33812127567254e-06,
+      "loss": 0.7318,
+      "step": 6201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6619288434644135,
+      "learning_rate": 6.335055848836702e-06,
+      "loss": 0.6514,
+      "step": 6202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.740794983691795,
+      "learning_rate": 6.3319908197732415e-06,
+      "loss": 0.6587,
+      "step": 6203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6781697662136037,
+      "learning_rate": 6.328926188814814e-06,
+      "loss": 0.8202,
+      "step": 6204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5546369098985757,
+      "learning_rate": 6.325861956294042e-06,
+      "loss": 0.6052,
+      "step": 6205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6583434658522769,
+      "learning_rate": 6.322798122543494e-06,
+      "loss": 0.6109,
+      "step": 6206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7530808501779143,
+      "learning_rate": 6.319734687895704e-06,
+      "loss": 0.6413,
+      "step": 6207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.611177983099102,
+      "learning_rate": 6.316671652683166e-06,
+      "loss": 0.7458,
+      "step": 6208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4791293149751368,
+      "learning_rate": 6.3136090172383136e-06,
+      "loss": 0.7304,
+      "step": 6209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7167625865519724,
+      "learning_rate": 6.310546781893556e-06,
+      "loss": 0.7791,
+      "step": 6210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6833743431663846,
+      "learning_rate": 6.307484946981251e-06,
+      "loss": 0.7106,
+      "step": 6211
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7007098813718895,
+      "learning_rate": 6.3044235128337065e-06,
+      "loss": 0.7332,
+      "step": 6212
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4874192642800728,
+      "learning_rate": 6.301362479783202e-06,
+      "loss": 0.648,
+      "step": 6213
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6123542295788824,
+      "learning_rate": 6.298301848161956e-06,
+      "loss": 0.6245,
+      "step": 6214
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4299700958218342,
+      "learning_rate": 6.295241618302156e-06,
+      "loss": 0.7062,
+      "step": 6215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6539641925588098,
+      "learning_rate": 6.292181790535947e-06,
+      "loss": 0.7223,
+      "step": 6216
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.778675414817209,
+      "learning_rate": 6.289122365195416e-06,
+      "loss": 0.8371,
+      "step": 6217
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.576626746519206,
+      "learning_rate": 6.286063342612625e-06,
+      "loss": 0.6435,
+      "step": 6218
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.514287004334272,
+      "learning_rate": 6.283004723119575e-06,
+      "loss": 0.6376,
+      "step": 6219
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5667375882028205,
+      "learning_rate": 6.2799465070482335e-06,
+      "loss": 0.7439,
+      "step": 6220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5966702971852884,
+      "learning_rate": 6.276888694730529e-06,
+      "loss": 0.674,
+      "step": 6221
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7417020646634092,
+      "learning_rate": 6.27383128649833e-06,
+      "loss": 0.7694,
+      "step": 6222
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.582188756240723,
+      "learning_rate": 6.270774282683476e-06,
+      "loss": 0.7764,
+      "step": 6223
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6387750113625243,
+      "learning_rate": 6.267717683617753e-06,
+      "loss": 0.7126,
+      "step": 6224
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7036047789699225,
+      "learning_rate": 6.26466148963291e-06,
+      "loss": 0.6611,
+      "step": 6225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4623154809803915,
+      "learning_rate": 6.261605701060649e-06,
+      "loss": 0.6599,
+      "step": 6226
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6301884041651713,
+      "learning_rate": 6.258550318232626e-06,
+      "loss": 0.6825,
+      "step": 6227
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.500453691932618,
+      "learning_rate": 6.255495341480455e-06,
+      "loss": 0.8056,
+      "step": 6228
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.861004993798108,
+      "learning_rate": 6.25244077113571e-06,
+      "loss": 0.6886,
+      "step": 6229
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4682486308246931,
+      "learning_rate": 6.249386607529914e-06,
+      "loss": 0.6634,
+      "step": 6230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5279737405454734,
+      "learning_rate": 6.246332850994547e-06,
+      "loss": 0.7362,
+      "step": 6231
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5951840003051774,
+      "learning_rate": 6.243279501861048e-06,
+      "loss": 0.6437,
+      "step": 6232
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5273387782005643,
+      "learning_rate": 6.240226560460811e-06,
+      "loss": 0.7581,
+      "step": 6233
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5694446493717522,
+      "learning_rate": 6.237174027125186e-06,
+      "loss": 0.7142,
+      "step": 6234
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7587644931185076,
+      "learning_rate": 6.234121902185475e-06,
+      "loss": 0.7101,
+      "step": 6235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.538172724798086,
+      "learning_rate": 6.231070185972943e-06,
+      "loss": 0.588,
+      "step": 6236
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.732688285618746,
+      "learning_rate": 6.2280188788187975e-06,
+      "loss": 0.5983,
+      "step": 6237
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.681520609984493,
+      "learning_rate": 6.224967981054216e-06,
+      "loss": 0.7516,
+      "step": 6238
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6385101464147946,
+      "learning_rate": 6.22191749301033e-06,
+      "loss": 0.7589,
+      "step": 6239
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7230564965309907,
+      "learning_rate": 6.218867415018213e-06,
+      "loss": 0.7187,
+      "step": 6240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6237010455135803,
+      "learning_rate": 6.215817747408912e-06,
+      "loss": 0.7512,
+      "step": 6241
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7284927765460951,
+      "learning_rate": 6.212768490513412e-06,
+      "loss": 0.7624,
+      "step": 6242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4408795649716721,
+      "learning_rate": 6.209719644662668e-06,
+      "loss": 0.6562,
+      "step": 6243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5648988173321112,
+      "learning_rate": 6.206671210187587e-06,
+      "loss": 0.6206,
+      "step": 6244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5329244582770034,
+      "learning_rate": 6.203623187419021e-06,
+      "loss": 0.7085,
+      "step": 6245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.689201951258469,
+      "learning_rate": 6.200575576687788e-06,
+      "loss": 0.6889,
+      "step": 6246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5068712420456922,
+      "learning_rate": 6.197528378324664e-06,
+      "loss": 0.6117,
+      "step": 6247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7289125382594037,
+      "learning_rate": 6.194481592660369e-06,
+      "loss": 0.7321,
+      "step": 6248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6075378730320198,
+      "learning_rate": 6.191435220025586e-06,
+      "loss": 0.7114,
+      "step": 6249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6979383319765728,
+      "learning_rate": 6.188389260750948e-06,
+      "loss": 0.6504,
+      "step": 6250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.741041947752599,
+      "learning_rate": 6.18534371516705e-06,
+      "loss": 0.7787,
+      "step": 6251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6692017023266297,
+      "learning_rate": 6.18229858360444e-06,
+      "loss": 0.7007,
+      "step": 6252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5863510304471515,
+      "learning_rate": 6.179253866393613e-06,
+      "loss": 0.7042,
+      "step": 6253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5770757934096424,
+      "learning_rate": 6.1762095638650336e-06,
+      "loss": 0.7247,
+      "step": 6254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5956129238205312,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.7633,
+      "step": 6255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6212786954144751,
+      "learning_rate": 6.170122204176194e-06,
+      "loss": 0.7317,
+      "step": 6256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6222304674140229,
+      "learning_rate": 6.167079147676632e-06,
+      "loss": 0.6724,
+      "step": 6257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5572506586899673,
+      "learning_rate": 6.164036507180684e-06,
+      "loss": 0.682,
+      "step": 6258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6638911977836073,
+      "learning_rate": 6.160994283018589e-06,
+      "loss": 0.7388,
+      "step": 6259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6200290901079435,
+      "learning_rate": 6.157952475520525e-06,
+      "loss": 0.7179,
+      "step": 6260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.775490359235819,
+      "learning_rate": 6.154911085016637e-06,
+      "loss": 0.6368,
+      "step": 6261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5992580931554452,
+      "learning_rate": 6.151870111837024e-06,
+      "loss": 0.8094,
+      "step": 6262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7821024117668556,
+      "learning_rate": 6.148829556311728e-06,
+      "loss": 0.8342,
+      "step": 6263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5027920068721097,
+      "learning_rate": 6.1457894187707644e-06,
+      "loss": 0.6714,
+      "step": 6264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7657574807299323,
+      "learning_rate": 6.142749699544079e-06,
+      "loss": 0.7252,
+      "step": 6265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5835231242874228,
+      "learning_rate": 6.1397103989615955e-06,
+      "loss": 0.6411,
+      "step": 6266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5892404817482912,
+      "learning_rate": 6.136671517353183e-06,
+      "loss": 0.7354,
+      "step": 6267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3784828776341291,
+      "learning_rate": 6.133633055048658e-06,
+      "loss": 0.5934,
+      "step": 6268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7207808030825036,
+      "learning_rate": 6.1305950123778024e-06,
+      "loss": 0.7213,
+      "step": 6269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5378430526726719,
+      "learning_rate": 6.127557389670351e-06,
+      "loss": 0.6262,
+      "step": 6270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5920669806506544,
+      "learning_rate": 6.124520187255983e-06,
+      "loss": 0.7416,
+      "step": 6271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.406596466187717,
+      "learning_rate": 6.121483405464346e-06,
+      "loss": 0.6638,
+      "step": 6272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7939125165954555,
+      "learning_rate": 6.118447044625027e-06,
+      "loss": 0.7147,
+      "step": 6273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.54819035552784,
+      "learning_rate": 6.115411105067582e-06,
+      "loss": 0.5916,
+      "step": 6274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7152824570643843,
+      "learning_rate": 6.112375587121518e-06,
+      "loss": 0.7127,
+      "step": 6275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7193281644819463,
+      "learning_rate": 6.109340491116284e-06,
+      "loss": 0.762,
+      "step": 6276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6424410423978122,
+      "learning_rate": 6.1063058173812975e-06,
+      "loss": 0.7889,
+      "step": 6277
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6949947318613092,
+      "learning_rate": 6.103271566245925e-06,
+      "loss": 0.7154,
+      "step": 6278
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.514080879501498,
+      "learning_rate": 6.1002377380394835e-06,
+      "loss": 0.7785,
+      "step": 6279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.506629183284391,
+      "learning_rate": 6.097204333091254e-06,
+      "loss": 0.7608,
+      "step": 6280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5244336613717824,
+      "learning_rate": 6.094171351730458e-06,
+      "loss": 0.7205,
+      "step": 6281
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.66852571528665,
+      "learning_rate": 6.0911387942862835e-06,
+      "loss": 0.7075,
+      "step": 6282
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6264908140490997,
+      "learning_rate": 6.0881066610878624e-06,
+      "loss": 0.7442,
+      "step": 6283
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4881575242339578,
+      "learning_rate": 6.08507495246429e-06,
+      "loss": 0.8112,
+      "step": 6284
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.810367992210238,
+      "learning_rate": 6.08204366874461e-06,
+      "loss": 0.8227,
+      "step": 6285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6755010174301157,
+      "learning_rate": 6.079012810257816e-06,
+      "loss": 0.7535,
+      "step": 6286
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5551086318305345,
+      "learning_rate": 6.075982377332868e-06,
+      "loss": 0.6764,
+      "step": 6287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5306851594326032,
+      "learning_rate": 6.072952370298667e-06,
+      "loss": 0.6971,
+      "step": 6288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7111329561521385,
+      "learning_rate": 6.069922789484073e-06,
+      "loss": 0.6651,
+      "step": 6289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6290435054738361,
+      "learning_rate": 6.066893635217903e-06,
+      "loss": 0.6927,
+      "step": 6290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.792078224753121,
+      "learning_rate": 6.0638649078289195e-06,
+      "loss": 0.7057,
+      "step": 6291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5726534950494397,
+      "learning_rate": 6.060836607645845e-06,
+      "loss": 0.7077,
+      "step": 6292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.414599565633372,
+      "learning_rate": 6.057808734997359e-06,
+      "loss": 0.6678,
+      "step": 6293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5506814668719011,
+      "learning_rate": 6.054781290212083e-06,
+      "loss": 0.617,
+      "step": 6294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5262725814759375,
+      "learning_rate": 6.051754273618605e-06,
+      "loss": 0.7637,
+      "step": 6295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5878570034248665,
+      "learning_rate": 6.0487276855454525e-06,
+      "loss": 0.7669,
+      "step": 6296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5719283837210194,
+      "learning_rate": 6.04570152632112e-06,
+      "loss": 0.7243,
+      "step": 6297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5875716436302951,
+      "learning_rate": 6.042675796274051e-06,
+      "loss": 0.6854,
+      "step": 6298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5012491845451224,
+      "learning_rate": 6.0396504957326365e-06,
+      "loss": 0.7472,
+      "step": 6299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5980477111010662,
+      "learning_rate": 6.03662562502523e-06,
+      "loss": 0.7216,
+      "step": 6300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5034426911917258,
+      "learning_rate": 6.033601184480129e-06,
+      "loss": 0.619,
+      "step": 6301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4686870305170632,
+      "learning_rate": 6.03057717442559e-06,
+      "loss": 0.638,
+      "step": 6302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5049175281810077,
+      "learning_rate": 6.02755359518983e-06,
+      "loss": 0.667,
+      "step": 6303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6873684104287792,
+      "learning_rate": 6.024530447101e-06,
+      "loss": 0.7237,
+      "step": 6304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7609432099220983,
+      "learning_rate": 6.021507730487221e-06,
+      "loss": 0.7489,
+      "step": 6305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4623537810787113,
+      "learning_rate": 6.018485445676565e-06,
+      "loss": 0.6506,
+      "step": 6306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.888703526424575,
+      "learning_rate": 6.015463592997046e-06,
+      "loss": 0.7954,
+      "step": 6307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6726747559744533,
+      "learning_rate": 6.0124421727766465e-06,
+      "loss": 0.7867,
+      "step": 6308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6863782626305706,
+      "learning_rate": 6.009421185343287e-06,
+      "loss": 0.7393,
+      "step": 6309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6499492821668567,
+      "learning_rate": 6.006400631024851e-06,
+      "loss": 0.8357,
+      "step": 6310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7095743812571649,
+      "learning_rate": 6.003380510149179e-06,
+      "loss": 0.6429,
+      "step": 6311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6484443288049906,
+      "learning_rate": 6.000360823044049e-06,
+      "loss": 0.7172,
+      "step": 6312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4579244853163298,
+      "learning_rate": 5.997341570037208e-06,
+      "loss": 0.729,
+      "step": 6313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4871202444789973,
+      "learning_rate": 5.994322751456339e-06,
+      "loss": 0.6165,
+      "step": 6314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5291037181344551,
+      "learning_rate": 5.9913043676290964e-06,
+      "loss": 0.718,
+      "step": 6315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5223742645046954,
+      "learning_rate": 5.988286418883078e-06,
+      "loss": 0.7463,
+      "step": 6316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3940083711431022,
+      "learning_rate": 5.985268905545829e-06,
+      "loss": 0.5375,
+      "step": 6317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6237783872688747,
+      "learning_rate": 5.98225182794486e-06,
+      "loss": 0.812,
+      "step": 6318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6176773287997215,
+      "learning_rate": 5.979235186407622e-06,
+      "loss": 0.8133,
+      "step": 6319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9042342799536476,
+      "learning_rate": 5.9762189812615254e-06,
+      "loss": 0.7468,
+      "step": 6320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7788211346001306,
+      "learning_rate": 5.973203212833939e-06,
+      "loss": 0.7365,
+      "step": 6321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8176366507210457,
+      "learning_rate": 5.970187881452168e-06,
+      "loss": 0.6977,
+      "step": 6322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4221286727959068,
+      "learning_rate": 5.967172987443482e-06,
+      "loss": 0.5771,
+      "step": 6323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4065211568246696,
+      "learning_rate": 5.964158531135106e-06,
+      "loss": 0.6418,
+      "step": 6324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.650739756578554,
+      "learning_rate": 5.961144512854205e-06,
+      "loss": 0.6653,
+      "step": 6325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6301746789006268,
+      "learning_rate": 5.958130932927908e-06,
+      "loss": 0.7417,
+      "step": 6326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4479738696607576,
+      "learning_rate": 5.955117791683289e-06,
+      "loss": 0.6565,
+      "step": 6327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5210492576094117,
+      "learning_rate": 5.952105089447378e-06,
+      "loss": 0.6512,
+      "step": 6328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.656886978316453,
+      "learning_rate": 5.94909282654716e-06,
+      "loss": 0.6448,
+      "step": 6329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6081903057337392,
+      "learning_rate": 5.946081003309565e-06,
+      "loss": 0.5581,
+      "step": 6330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.753778483406041,
+      "learning_rate": 5.943069620061481e-06,
+      "loss": 0.7395,
+      "step": 6331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6802658251468272,
+      "learning_rate": 5.940058677129748e-06,
+      "loss": 0.6784,
+      "step": 6332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6891268164739461,
+      "learning_rate": 5.937048174841153e-06,
+      "loss": 0.7862,
+      "step": 6333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.918030920602363,
+      "learning_rate": 5.934038113522442e-06,
+      "loss": 0.7078,
+      "step": 6334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7695806628166773,
+      "learning_rate": 5.9310284935003106e-06,
+      "loss": 0.7207,
+      "step": 6335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5286937983901505,
+      "learning_rate": 5.928019315101403e-06,
+      "loss": 0.639,
+      "step": 6336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5447517041351149,
+      "learning_rate": 5.92501057865232e-06,
+      "loss": 0.6319,
+      "step": 6337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5609798103338413,
+      "learning_rate": 5.922002284479614e-06,
+      "loss": 0.642,
+      "step": 6338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5687889938638144,
+      "learning_rate": 5.9189944329097885e-06,
+      "loss": 0.7272,
+      "step": 6339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6224286950572846,
+      "learning_rate": 5.915987024269294e-06,
+      "loss": 0.7058,
+      "step": 6340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5027567146457037,
+      "learning_rate": 5.912980058884543e-06,
+      "loss": 0.7581,
+      "step": 6341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7082554644942154,
+      "learning_rate": 5.909973537081893e-06,
+      "loss": 0.7154,
+      "step": 6342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.691328540293944,
+      "learning_rate": 5.9069674591876535e-06,
+      "loss": 0.7977,
+      "step": 6343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6256752787523274,
+      "learning_rate": 5.903961825528092e-06,
+      "loss": 0.6784,
+      "step": 6344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.556542188553163,
+      "learning_rate": 5.900956636429416e-06,
+      "loss": 0.6217,
+      "step": 6345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7067246350845486,
+      "learning_rate": 5.8979518922177956e-06,
+      "loss": 0.7909,
+      "step": 6346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5820138550985727,
+      "learning_rate": 5.8949475932193505e-06,
+      "loss": 0.6841,
+      "step": 6347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.540678213252206,
+      "learning_rate": 5.891943739760144e-06,
+      "loss": 0.6931,
+      "step": 6348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6771935590518536,
+      "learning_rate": 5.888940332166209e-06,
+      "loss": 0.6764,
+      "step": 6349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7257977664771067,
+      "learning_rate": 5.885937370763503e-06,
+      "loss": 0.6693,
+      "step": 6350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7398417630894818,
+      "learning_rate": 5.882934855877962e-06,
+      "loss": 0.7415,
+      "step": 6351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8211802165336226,
+      "learning_rate": 5.8799327878354615e-06,
+      "loss": 0.7329,
+      "step": 6352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6342677687069151,
+      "learning_rate": 5.876931166961823e-06,
+      "loss": 0.758,
+      "step": 6353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5043393902621687,
+      "learning_rate": 5.873929993582832e-06,
+      "loss": 0.6999,
+      "step": 6354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6347981025495912,
+      "learning_rate": 5.8709292680242114e-06,
+      "loss": 0.6204,
+      "step": 6355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5008672609314513,
+      "learning_rate": 5.867928990611647e-06,
+      "loss": 0.5959,
+      "step": 6356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6378971953945671,
+      "learning_rate": 5.864929161670778e-06,
+      "loss": 0.7179,
+      "step": 6357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5754498848685459,
+      "learning_rate": 5.861929781527178e-06,
+      "loss": 0.698,
+      "step": 6358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.553793198749774,
+      "learning_rate": 5.858930850506388e-06,
+      "loss": 0.6156,
+      "step": 6359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.458805236794188,
+      "learning_rate": 5.8559323689338985e-06,
+      "loss": 0.7464,
+      "step": 6360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5183497123812193,
+      "learning_rate": 5.852934337135142e-06,
+      "loss": 0.6604,
+      "step": 6361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7357794415167893,
+      "learning_rate": 5.849936755435513e-06,
+      "loss": 0.75,
+      "step": 6362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4167584344055437,
+      "learning_rate": 5.846939624160346e-06,
+      "loss": 0.6739,
+      "step": 6363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5571878965670656,
+      "learning_rate": 5.843942943634937e-06,
+      "loss": 0.7643,
+      "step": 6364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.674640396827517,
+      "learning_rate": 5.840946714184531e-06,
+      "loss": 0.8063,
+      "step": 6365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6494268241629362,
+      "learning_rate": 5.8379509361343154e-06,
+      "loss": 0.6854,
+      "step": 6366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5840646906041485,
+      "learning_rate": 5.834955609809443e-06,
+      "loss": 0.6196,
+      "step": 6367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6550653534082749,
+      "learning_rate": 5.831960735534999e-06,
+      "loss": 0.7491,
+      "step": 6368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7931312062298828,
+      "learning_rate": 5.8289663136360376e-06,
+      "loss": 0.7552,
+      "step": 6369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6495599365387563,
+      "learning_rate": 5.8259723444375605e-06,
+      "loss": 0.805,
+      "step": 6370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6737151146352796,
+      "learning_rate": 5.822978828264505e-06,
+      "loss": 0.7276,
+      "step": 6371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6837966851701676,
+      "learning_rate": 5.819985765441781e-06,
+      "loss": 0.7014,
+      "step": 6372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5072843482812206,
+      "learning_rate": 5.81699315629423e-06,
+      "loss": 0.6911,
+      "step": 6373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7206761150182308,
+      "learning_rate": 5.814001001146657e-06,
+      "loss": 0.6909,
+      "step": 6374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5046183305177292,
+      "learning_rate": 5.8110093003238175e-06,
+      "loss": 0.627,
+      "step": 6375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4910705578526564,
+      "learning_rate": 5.808018054150406e-06,
+      "loss": 0.6718,
+      "step": 6376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5485078588939911,
+      "learning_rate": 5.805027262951079e-06,
+      "loss": 0.7327,
+      "step": 6377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6506032689070136,
+      "learning_rate": 5.802036927050447e-06,
+      "loss": 0.6427,
+      "step": 6378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6698136293154233,
+      "learning_rate": 5.799047046773052e-06,
+      "loss": 0.8641,
+      "step": 6379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5663409493127316,
+      "learning_rate": 5.7960576224434074e-06,
+      "loss": 0.6532,
+      "step": 6380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.485427935130718,
+      "learning_rate": 5.793068654385963e-06,
+      "loss": 0.6535,
+      "step": 6381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6252387841757945,
+      "learning_rate": 5.790080142925128e-06,
+      "loss": 0.7337,
+      "step": 6382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6319477125343451,
+      "learning_rate": 5.7870920883852595e-06,
+      "loss": 0.6668,
+      "step": 6383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8594213711478256,
+      "learning_rate": 5.7841044910906585e-06,
+      "loss": 0.7431,
+      "step": 6384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6592215105240347,
+      "learning_rate": 5.781117351365592e-06,
+      "loss": 0.7135,
+      "step": 6385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.517474499115124,
+      "learning_rate": 5.778130669534254e-06,
+      "loss": 0.6853,
+      "step": 6386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4892996147591693,
+      "learning_rate": 5.775144445920811e-06,
+      "loss": 0.7471,
+      "step": 6387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5256135918516867,
+      "learning_rate": 5.772158680849374e-06,
+      "loss": 0.7412,
+      "step": 6388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5529282088524847,
+      "learning_rate": 5.769173374643991e-06,
+      "loss": 0.6539,
+      "step": 6389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6485522629940237,
+      "learning_rate": 5.766188527628679e-06,
+      "loss": 0.7233,
+      "step": 6390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6621024165012506,
+      "learning_rate": 5.76320414012739e-06,
+      "loss": 0.6734,
+      "step": 6391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7848678387530656,
+      "learning_rate": 5.760220212464034e-06,
+      "loss": 0.7261,
+      "step": 6392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4887429010798594,
+      "learning_rate": 5.757236744962476e-06,
+      "loss": 0.6584,
+      "step": 6393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.67450143515445,
+      "learning_rate": 5.754253737946516e-06,
+      "loss": 0.6852,
+      "step": 6394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5312323954574496,
+      "learning_rate": 5.751271191739917e-06,
+      "loss": 0.7058,
+      "step": 6395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6236234571249322,
+      "learning_rate": 5.748289106666392e-06,
+      "loss": 0.7407,
+      "step": 6396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6408794958238797,
+      "learning_rate": 5.7453074830495896e-06,
+      "loss": 0.7075,
+      "step": 6397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9122164979695884,
+      "learning_rate": 5.742326321213127e-06,
+      "loss": 0.6851,
+      "step": 6398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5172361668955892,
+      "learning_rate": 5.739345621480559e-06,
+      "loss": 0.6739,
+      "step": 6399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5217073707337485,
+      "learning_rate": 5.736365384175393e-06,
+      "loss": 0.5661,
+      "step": 6400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.681497755371146,
+      "learning_rate": 5.733385609621092e-06,
+      "loss": 0.7032,
+      "step": 6401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4642727363740708,
+      "learning_rate": 5.730406298141058e-06,
+      "loss": 0.6612,
+      "step": 6402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6701836113853448,
+      "learning_rate": 5.7274274500586535e-06,
+      "loss": 0.6383,
+      "step": 6403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5146756310866385,
+      "learning_rate": 5.724449065697182e-06,
+      "loss": 0.6371,
+      "step": 6404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5455245536838689,
+      "learning_rate": 5.721471145379901e-06,
+      "loss": 0.6639,
+      "step": 6405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.705302002513825,
+      "learning_rate": 5.718493689430022e-06,
+      "loss": 0.807,
+      "step": 6406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7491318011537786,
+      "learning_rate": 5.715516698170696e-06,
+      "loss": 0.8477,
+      "step": 6407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.666722548545367,
+      "learning_rate": 5.712540171925029e-06,
+      "loss": 0.7674,
+      "step": 6408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7357789038610587,
+      "learning_rate": 5.709564111016081e-06,
+      "loss": 0.6924,
+      "step": 6409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6110366786415917,
+      "learning_rate": 5.706588515766851e-06,
+      "loss": 0.6787,
+      "step": 6410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6350931099155135,
+      "learning_rate": 5.7036133865003e-06,
+      "loss": 0.698,
+      "step": 6411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7228823458815403,
+      "learning_rate": 5.700638723539325e-06,
+      "loss": 0.7233,
+      "step": 6412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6021431264605037,
+      "learning_rate": 5.69766452720678e-06,
+      "loss": 0.6322,
+      "step": 6413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6334750931777735,
+      "learning_rate": 5.694690797825475e-06,
+      "loss": 0.6427,
+      "step": 6414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5833645717447067,
+      "learning_rate": 5.691717535718151e-06,
+      "loss": 0.7319,
+      "step": 6415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7498137753102825,
+      "learning_rate": 5.688744741207516e-06,
+      "loss": 0.6878,
+      "step": 6416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5676021784469838,
+      "learning_rate": 5.6857724146162215e-06,
+      "loss": 0.6379,
+      "step": 6417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5217944573670767,
+      "learning_rate": 5.682800556266862e-06,
+      "loss": 0.6389,
+      "step": 6418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3689128575405203,
+      "learning_rate": 5.6798291664819875e-06,
+      "loss": 0.5964,
+      "step": 6419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5680466041697287,
+      "learning_rate": 5.676858245584103e-06,
+      "loss": 0.6202,
+      "step": 6420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.755950732711214,
+      "learning_rate": 5.6738877938956426e-06,
+      "loss": 0.7522,
+      "step": 6421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7249538817905252,
+      "learning_rate": 5.6709178117390105e-06,
+      "loss": 0.7686,
+      "step": 6422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.69730402680779,
+      "learning_rate": 5.667948299436555e-06,
+      "loss": 0.682,
+      "step": 6423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6589657016064634,
+      "learning_rate": 5.6649792573105625e-06,
+      "loss": 0.6969,
+      "step": 6424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.695822779042538,
+      "learning_rate": 5.662010685683279e-06,
+      "loss": 0.6876,
+      "step": 6425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7421322612105266,
+      "learning_rate": 5.6590425848769e-06,
+      "loss": 0.6974,
+      "step": 6426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8334794155879066,
+      "learning_rate": 5.6560749552135605e-06,
+      "loss": 0.8303,
+      "step": 6427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6470435016775014,
+      "learning_rate": 5.653107797015354e-06,
+      "loss": 0.6621,
+      "step": 6428
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.627521191505139,
+      "learning_rate": 5.6501411106043205e-06,
+      "loss": 0.6695,
+      "step": 6429
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5213801668421556,
+      "learning_rate": 5.647174896302442e-06,
+      "loss": 0.6911,
+      "step": 6430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5363389440683215,
+      "learning_rate": 5.644209154431662e-06,
+      "loss": 0.6973,
+      "step": 6431
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4381990884890545,
+      "learning_rate": 5.641243885313856e-06,
+      "loss": 0.697,
+      "step": 6432
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7122163825991126,
+      "learning_rate": 5.6382790892708665e-06,
+      "loss": 0.6942,
+      "step": 6433
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5466831181701948,
+      "learning_rate": 5.635314766624474e-06,
+      "loss": 0.6505,
+      "step": 6434
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.543566320891912,
+      "learning_rate": 5.632350917696402e-06,
+      "loss": 0.6747,
+      "step": 6435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6487087786876302,
+      "learning_rate": 5.629387542808338e-06,
+      "loss": 0.6439,
+      "step": 6436
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5662196712883276,
+      "learning_rate": 5.626424642281909e-06,
+      "loss": 0.6167,
+      "step": 6437
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6125241640825956,
+      "learning_rate": 5.623462216438689e-06,
+      "loss": 0.6974,
+      "step": 6438
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4991815498617118,
+      "learning_rate": 5.620500265600206e-06,
+      "loss": 0.7292,
+      "step": 6439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6384111399872088,
+      "learning_rate": 5.617538790087927e-06,
+      "loss": 0.653,
+      "step": 6440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3135639297560717,
+      "learning_rate": 5.614577790223279e-06,
+      "loss": 0.61,
+      "step": 6441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5333413114465992,
+      "learning_rate": 5.611617266327636e-06,
+      "loss": 0.6731,
+      "step": 6442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6331806993463807,
+      "learning_rate": 5.608657218722309e-06,
+      "loss": 0.7274,
+      "step": 6443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6925118285778062,
+      "learning_rate": 5.6056976477285695e-06,
+      "loss": 0.884,
+      "step": 6444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4276036664358116,
+      "learning_rate": 5.602738553667629e-06,
+      "loss": 0.6985,
+      "step": 6445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5209224336151763,
+      "learning_rate": 5.599779936860652e-06,
+      "loss": 0.6954,
+      "step": 6446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.611923903199948,
+      "learning_rate": 5.5968217976287565e-06,
+      "loss": 0.6656,
+      "step": 6447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8579107068263057,
+      "learning_rate": 5.593864136292992e-06,
+      "loss": 0.707,
+      "step": 6448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5452524877049152,
+      "learning_rate": 5.5909069531743755e-06,
+      "loss": 0.6808,
+      "step": 6449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.698242540548482,
+      "learning_rate": 5.5879502485938544e-06,
+      "loss": 0.7089,
+      "step": 6450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8162676945644611,
+      "learning_rate": 5.584994022872337e-06,
+      "loss": 0.7118,
+      "step": 6451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.538992476391323,
+      "learning_rate": 5.582038276330679e-06,
+      "loss": 0.7687,
+      "step": 6452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7325920782664779,
+      "learning_rate": 5.5790830092896744e-06,
+      "loss": 0.6782,
+      "step": 6453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7047326937943428,
+      "learning_rate": 5.576128222070072e-06,
+      "loss": 0.7279,
+      "step": 6454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5326744108994126,
+      "learning_rate": 5.573173914992575e-06,
+      "loss": 0.7136,
+      "step": 6455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5658639914644978,
+      "learning_rate": 5.570220088377817e-06,
+      "loss": 0.6877,
+      "step": 6456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5407873449624567,
+      "learning_rate": 5.567266742546398e-06,
+      "loss": 0.7392,
+      "step": 6457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7165329737166082,
+      "learning_rate": 5.56431387781885e-06,
+      "loss": 0.6999,
+      "step": 6458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5792158050361675,
+      "learning_rate": 5.5613614945156644e-06,
+      "loss": 0.8257,
+      "step": 6459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6861064784060478,
+      "learning_rate": 5.558409592957281e-06,
+      "loss": 0.6937,
+      "step": 6460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6622203408620764,
+      "learning_rate": 5.555458173464073e-06,
+      "loss": 0.7079,
+      "step": 6461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6179041108888554,
+      "learning_rate": 5.552507236356378e-06,
+      "loss": 0.6671,
+      "step": 6462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5561973558428366,
+      "learning_rate": 5.549556781954468e-06,
+      "loss": 0.688,
+      "step": 6463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8770461530985756,
+      "learning_rate": 5.546606810578574e-06,
+      "loss": 0.7056,
+      "step": 6464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6639773368960784,
+      "learning_rate": 5.54365732254887e-06,
+      "loss": 0.8691,
+      "step": 6465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.766616000861856,
+      "learning_rate": 5.540708318185471e-06,
+      "loss": 0.6345,
+      "step": 6466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7447860975157186,
+      "learning_rate": 5.537759797808451e-06,
+      "loss": 0.8052,
+      "step": 6467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6314521492158323,
+      "learning_rate": 5.5348117617378205e-06,
+      "loss": 0.7425,
+      "step": 6468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6169731780187688,
+      "learning_rate": 5.531864210293545e-06,
+      "loss": 0.6888,
+      "step": 6469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8868048761897305,
+      "learning_rate": 5.52891714379554e-06,
+      "loss": 0.6861,
+      "step": 6470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.563665870685886,
+      "learning_rate": 5.525970562563656e-06,
+      "loss": 0.7059,
+      "step": 6471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.66926400990454,
+      "learning_rate": 5.5230244669177e-06,
+      "loss": 0.7607,
+      "step": 6472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6671494780823763,
+      "learning_rate": 5.520078857177431e-06,
+      "loss": 0.7621,
+      "step": 6473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.574392066987864,
+      "learning_rate": 5.5171337336625385e-06,
+      "loss": 0.718,
+      "step": 6474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.77239094210856,
+      "learning_rate": 5.5141890966926814e-06,
+      "loss": 0.7875,
+      "step": 6475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.423741447571245,
+      "learning_rate": 5.511244946587442e-06,
+      "loss": 0.7009,
+      "step": 6476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.555072089538958,
+      "learning_rate": 5.5083012836663675e-06,
+      "loss": 0.6933,
+      "step": 6477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.685968478521124,
+      "learning_rate": 5.5053581082489505e-06,
+      "loss": 0.6945,
+      "step": 6478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4061025474606934,
+      "learning_rate": 5.502415420654619e-06,
+      "loss": 0.5818,
+      "step": 6479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6721597452098804,
+      "learning_rate": 5.4994732212027625e-06,
+      "loss": 0.7269,
+      "step": 6480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6885123796718169,
+      "learning_rate": 5.4965315102127035e-06,
+      "loss": 0.7111,
+      "step": 6481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.587768016336439,
+      "learning_rate": 5.4935902880037224e-06,
+      "loss": 0.7552,
+      "step": 6482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6544396399034218,
+      "learning_rate": 5.490649554895047e-06,
+      "loss": 0.7939,
+      "step": 6483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.903730177094068,
+      "learning_rate": 5.487709311205842e-06,
+      "loss": 0.703,
+      "step": 6484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5464431344845322,
+      "learning_rate": 5.48476955725523e-06,
+      "loss": 0.6926,
+      "step": 6485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6445360027395115,
+      "learning_rate": 5.481830293362267e-06,
+      "loss": 0.72,
+      "step": 6486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4844153552483355,
+      "learning_rate": 5.47889151984597e-06,
+      "loss": 0.6338,
+      "step": 6487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7527504474138509,
+      "learning_rate": 5.475953237025301e-06,
+      "loss": 0.7561,
+      "step": 6488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5552778353388261,
+      "learning_rate": 5.473015445219155e-06,
+      "loss": 0.821,
+      "step": 6489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6307048291861013,
+      "learning_rate": 5.470078144746389e-06,
+      "loss": 0.7274,
+      "step": 6490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7001117892238,
+      "learning_rate": 5.467141335925804e-06,
+      "loss": 0.7728,
+      "step": 6491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5869091692069035,
+      "learning_rate": 5.464205019076138e-06,
+      "loss": 0.6786,
+      "step": 6492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6329469565910237,
+      "learning_rate": 5.461269194516088e-06,
+      "loss": 0.7075,
+      "step": 6493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8988642438158536,
+      "learning_rate": 5.458333862564286e-06,
+      "loss": 0.7586,
+      "step": 6494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6633595330570712,
+      "learning_rate": 5.4553990235393184e-06,
+      "loss": 0.6541,
+      "step": 6495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5781666882673357,
+      "learning_rate": 5.452464677759721e-06,
+      "loss": 0.6325,
+      "step": 6496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5661497583477781,
+      "learning_rate": 5.449530825543965e-06,
+      "loss": 0.7228,
+      "step": 6497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5866430847198567,
+      "learning_rate": 5.446597467210479e-06,
+      "loss": 0.5912,
+      "step": 6498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6481833752296238,
+      "learning_rate": 5.443664603077628e-06,
+      "loss": 0.8418,
+      "step": 6499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.695607932716036,
+      "learning_rate": 5.44073223346373e-06,
+      "loss": 0.7262,
+      "step": 6500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5218003829304017,
+      "learning_rate": 5.437800358687054e-06,
+      "loss": 0.7119,
+      "step": 6501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5239834220771615,
+      "learning_rate": 5.4348689790658006e-06,
+      "loss": 0.6818,
+      "step": 6502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5960102475372329,
+      "learning_rate": 5.431938094918132e-06,
+      "loss": 0.7411,
+      "step": 6503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.657548516507031,
+      "learning_rate": 5.4290077065621426e-06,
+      "loss": 0.7023,
+      "step": 6504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6262940709200162,
+      "learning_rate": 5.426077814315885e-06,
+      "loss": 0.7244,
+      "step": 6505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6390654376483609,
+      "learning_rate": 5.423148418497356e-06,
+      "loss": 0.5825,
+      "step": 6506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6031873520521178,
+      "learning_rate": 5.4202195194244876e-06,
+      "loss": 0.7294,
+      "step": 6507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7798635128022935,
+      "learning_rate": 5.417291117415172e-06,
+      "loss": 0.6766,
+      "step": 6508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6385895753692539,
+      "learning_rate": 5.414363212787244e-06,
+      "loss": 0.6456,
+      "step": 6509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.586186792105555,
+      "learning_rate": 5.411435805858473e-06,
+      "loss": 0.7585,
+      "step": 6510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6013166183073786,
+      "learning_rate": 5.4085088969465916e-06,
+      "loss": 0.703,
+      "step": 6511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5405890246283256,
+      "learning_rate": 5.405582486369264e-06,
+      "loss": 0.6945,
+      "step": 6512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4703347207965698,
+      "learning_rate": 5.402656574444108e-06,
+      "loss": 0.709,
+      "step": 6513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7235942090586773,
+      "learning_rate": 5.399731161488691e-06,
+      "loss": 0.7877,
+      "step": 6514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.576014597382119,
+      "learning_rate": 5.3968062478205115e-06,
+      "loss": 0.705,
+      "step": 6515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4943026192192757,
+      "learning_rate": 5.393881833757033e-06,
+      "loss": 0.6441,
+      "step": 6516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4372588715266486,
+      "learning_rate": 5.390957919615645e-06,
+      "loss": 0.6064,
+      "step": 6517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5746568492901751,
+      "learning_rate": 5.388034505713697e-06,
+      "loss": 0.7424,
+      "step": 6518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6185110839386305,
+      "learning_rate": 5.385111592368486e-06,
+      "loss": 0.6514,
+      "step": 6519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.795299732229972,
+      "learning_rate": 5.382189179897237e-06,
+      "loss": 0.7573,
+      "step": 6520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6681414848347083,
+      "learning_rate": 5.379267268617139e-06,
+      "loss": 0.731,
+      "step": 6521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8258298354662874,
+      "learning_rate": 5.376345858845322e-06,
+      "loss": 0.7484,
+      "step": 6522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4220644902625972,
+      "learning_rate": 5.373424950898854e-06,
+      "loss": 0.5895,
+      "step": 6523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5616930856307956,
+      "learning_rate": 5.3705045450947546e-06,
+      "loss": 0.7315,
+      "step": 6524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.650050715075898,
+      "learning_rate": 5.367584641749994e-06,
+      "loss": 0.7211,
+      "step": 6525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5675250874675495,
+      "learning_rate": 5.3646652411814735e-06,
+      "loss": 0.767,
+      "step": 6526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7880041175999453,
+      "learning_rate": 5.361746343706053e-06,
+      "loss": 0.7265,
+      "step": 6527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4506290455380577,
+      "learning_rate": 5.358827949640536e-06,
+      "loss": 0.7461,
+      "step": 6528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9009951006619712,
+      "learning_rate": 5.3559100593016615e-06,
+      "loss": 0.6162,
+      "step": 6529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6192604915096651,
+      "learning_rate": 5.352992673006125e-06,
+      "loss": 0.6748,
+      "step": 6530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6344829669114818,
+      "learning_rate": 5.350075791070566e-06,
+      "loss": 0.7157,
+      "step": 6531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7282190505058468,
+      "learning_rate": 5.34715941381156e-06,
+      "loss": 0.7053,
+      "step": 6532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5616856047012901,
+      "learning_rate": 5.344243541545637e-06,
+      "loss": 0.7304,
+      "step": 6533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.488895832298304,
+      "learning_rate": 5.341328174589273e-06,
+      "loss": 0.7114,
+      "step": 6534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7697051118099179,
+      "learning_rate": 5.338413313258879e-06,
+      "loss": 0.7133,
+      "step": 6535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6997375561691037,
+      "learning_rate": 5.335498957870821e-06,
+      "loss": 0.7103,
+      "step": 6536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.737073289454585,
+      "learning_rate": 5.332585108741411e-06,
+      "loss": 0.7116,
+      "step": 6537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.69707998251365,
+      "learning_rate": 5.329671766186892e-06,
+      "loss": 0.6982,
+      "step": 6538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5938705443800656,
+      "learning_rate": 5.32675893052347e-06,
+      "loss": 0.7617,
+      "step": 6539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6785160562909558,
+      "learning_rate": 5.323846602067283e-06,
+      "loss": 0.7133,
+      "step": 6540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7254541858211765,
+      "learning_rate": 5.320934781134419e-06,
+      "loss": 0.7006,
+      "step": 6541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4717849939657723,
+      "learning_rate": 5.318023468040918e-06,
+      "loss": 0.6648,
+      "step": 6542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.542910712759783,
+      "learning_rate": 5.315112663102746e-06,
+      "loss": 0.7322,
+      "step": 6543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.737205889574892,
+      "learning_rate": 5.312202366635831e-06,
+      "loss": 0.7091,
+      "step": 6544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5309583554809463,
+      "learning_rate": 5.3092925789560445e-06,
+      "loss": 0.6099,
+      "step": 6545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.480662378918754,
+      "learning_rate": 5.30638330037919e-06,
+      "loss": 0.749,
+      "step": 6546
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.461291497402869,
+      "learning_rate": 5.303474531221033e-06,
+      "loss": 0.7021,
+      "step": 6547
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.65880915293672,
+      "learning_rate": 5.300566271797265e-06,
+      "loss": 0.6811,
+      "step": 6548
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3833613813366754,
+      "learning_rate": 5.297658522423539e-06,
+      "loss": 0.65,
+      "step": 6549
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.587303853860589,
+      "learning_rate": 5.294751283415446e-06,
+      "loss": 0.76,
+      "step": 6550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.563782664562314,
+      "learning_rate": 5.2918445550885165e-06,
+      "loss": 0.6299,
+      "step": 6551
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4450028756994537,
+      "learning_rate": 5.288938337758236e-06,
+      "loss": 0.5837,
+      "step": 6552
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.663183818495271,
+      "learning_rate": 5.286032631740023e-06,
+      "loss": 0.7403,
+      "step": 6553
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.491077128099564,
+      "learning_rate": 5.28312743734925e-06,
+      "loss": 0.6107,
+      "step": 6554
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7013609398124023,
+      "learning_rate": 5.280222754901234e-06,
+      "loss": 0.7547,
+      "step": 6555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.488129397639651,
+      "learning_rate": 5.277318584711224e-06,
+      "loss": 0.5939,
+      "step": 6556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6660977197771891,
+      "learning_rate": 5.274414927094433e-06,
+      "loss": 0.718,
+      "step": 6557
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.596142506255152,
+      "learning_rate": 5.271511782365997e-06,
+      "loss": 0.5755,
+      "step": 6558
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.570684630382438,
+      "learning_rate": 5.268609150841011e-06,
+      "loss": 0.7117,
+      "step": 6559
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3862051356848353,
+      "learning_rate": 5.265707032834515e-06,
+      "loss": 0.8393,
+      "step": 6560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.676536115535365,
+      "learning_rate": 5.262805428661481e-06,
+      "loss": 0.8902,
+      "step": 6561
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6601852418073246,
+      "learning_rate": 5.259904338636839e-06,
+      "loss": 0.6665,
+      "step": 6562
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7040187006990393,
+      "learning_rate": 5.257003763075452e-06,
+      "loss": 0.7277,
+      "step": 6563
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.683236229240097,
+      "learning_rate": 5.2541037022921335e-06,
+      "loss": 0.7488,
+      "step": 6564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4747881576506794,
+      "learning_rate": 5.251204156601645e-06,
+      "loss": 0.6613,
+      "step": 6565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6054185824390816,
+      "learning_rate": 5.248305126318678e-06,
+      "loss": 0.7666,
+      "step": 6566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4386913476814391,
+      "learning_rate": 5.245406611757882e-06,
+      "loss": 0.7446,
+      "step": 6567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5303464616012228,
+      "learning_rate": 5.242508613233848e-06,
+      "loss": 0.6642,
+      "step": 6568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6563424630765726,
+      "learning_rate": 5.239611131061102e-06,
+      "loss": 0.6124,
+      "step": 6569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4331120110368254,
+      "learning_rate": 5.236714165554126e-06,
+      "loss": 0.6598,
+      "step": 6570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.298885302331489,
+      "learning_rate": 5.233817717027337e-06,
+      "loss": 0.6008,
+      "step": 6571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.603739287059268,
+      "learning_rate": 5.230921785795097e-06,
+      "loss": 0.6126,
+      "step": 6572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5185383591183974,
+      "learning_rate": 5.228026372171722e-06,
+      "loss": 0.6423,
+      "step": 6573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.408739492333563,
+      "learning_rate": 5.225131476471455e-06,
+      "loss": 0.6603,
+      "step": 6574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5087200124354232,
+      "learning_rate": 5.2222370990085006e-06,
+      "loss": 0.5578,
+      "step": 6575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6882350770849142,
+      "learning_rate": 5.219343240096988e-06,
+      "loss": 0.7846,
+      "step": 6576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4165427103087591,
+      "learning_rate": 5.216449900051006e-06,
+      "loss": 0.6297,
+      "step": 6577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6031125511277677,
+      "learning_rate": 5.213557079184584e-06,
+      "loss": 0.6257,
+      "step": 6578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6330463089764724,
+      "learning_rate": 5.210664777811687e-06,
+      "loss": 0.7085,
+      "step": 6579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.55459047365538,
+      "learning_rate": 5.207772996246234e-06,
+      "loss": 0.718,
+      "step": 6580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8562593153317937,
+      "learning_rate": 5.204881734802078e-06,
+      "loss": 0.7194,
+      "step": 6581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6494287308768552,
+      "learning_rate": 5.201990993793022e-06,
+      "loss": 0.7813,
+      "step": 6582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6253578733828447,
+      "learning_rate": 5.199100773532815e-06,
+      "loss": 0.7396,
+      "step": 6583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7042981931473753,
+      "learning_rate": 5.196211074335136e-06,
+      "loss": 0.7305,
+      "step": 6584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4128951522222983,
+      "learning_rate": 5.1933218965136235e-06,
+      "loss": 0.7753,
+      "step": 6585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6599676572854145,
+      "learning_rate": 5.190433240381855e-06,
+      "loss": 0.7574,
+      "step": 6586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6950322153257795,
+      "learning_rate": 5.18754510625334e-06,
+      "loss": 0.6192,
+      "step": 6587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5876474049189262,
+      "learning_rate": 5.1846574944415505e-06,
+      "loss": 0.6778,
+      "step": 6588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7067641494328831,
+      "learning_rate": 5.181770405259883e-06,
+      "loss": 0.6738,
+      "step": 6589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.560063028944311,
+      "learning_rate": 5.1788838390216874e-06,
+      "loss": 0.604,
+      "step": 6590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6402935150003717,
+      "learning_rate": 5.175997796040262e-06,
+      "loss": 0.7236,
+      "step": 6591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4941234108482795,
+      "learning_rate": 5.173112276628834e-06,
+      "loss": 0.8053,
+      "step": 6592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8240564355995594,
+      "learning_rate": 5.170227281100587e-06,
+      "loss": 0.6713,
+      "step": 6593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5962204639297288,
+      "learning_rate": 5.167342809768636e-06,
+      "loss": 0.7,
+      "step": 6594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6994359482698425,
+      "learning_rate": 5.164458862946049e-06,
+      "loss": 0.6477,
+      "step": 6595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5502079375142361,
+      "learning_rate": 5.1615754409458365e-06,
+      "loss": 0.628,
+      "step": 6596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7346230320590885,
+      "learning_rate": 5.158692544080941e-06,
+      "loss": 0.7447,
+      "step": 6597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4492600740682662,
+      "learning_rate": 5.155810172664264e-06,
+      "loss": 0.6451,
+      "step": 6598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6226120765657546,
+      "learning_rate": 5.152928327008635e-06,
+      "loss": 0.5691,
+      "step": 6599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6380828656883044,
+      "learning_rate": 5.1500470074268375e-06,
+      "loss": 0.6748,
+      "step": 6600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5436601441971176,
+      "learning_rate": 5.147166214231595e-06,
+      "loss": 0.6615,
+      "step": 6601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4598286166313794,
+      "learning_rate": 5.144285947735566e-06,
+      "loss": 0.6504,
+      "step": 6602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.691712192890952,
+      "learning_rate": 5.141406208251363e-06,
+      "loss": 0.6155,
+      "step": 6603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4712765944949058,
+      "learning_rate": 5.13852699609154e-06,
+      "loss": 0.6557,
+      "step": 6604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6182224948885766,
+      "learning_rate": 5.135648311568583e-06,
+      "loss": 0.7785,
+      "step": 6605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7863311346040405,
+      "learning_rate": 5.132770154994935e-06,
+      "loss": 0.7662,
+      "step": 6606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6875108524698368,
+      "learning_rate": 5.1298925266829694e-06,
+      "loss": 0.681,
+      "step": 6607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6087828341371864,
+      "learning_rate": 5.127015426945008e-06,
+      "loss": 0.6428,
+      "step": 6608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4344497230237643,
+      "learning_rate": 5.124138856093321e-06,
+      "loss": 0.6741,
+      "step": 6609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.630878458485287,
+      "learning_rate": 5.1212628144401086e-06,
+      "loss": 0.7074,
+      "step": 6610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5157406026220226,
+      "learning_rate": 5.118387302297527e-06,
+      "loss": 0.674,
+      "step": 6611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7480585578015724,
+      "learning_rate": 5.115512319977659e-06,
+      "loss": 0.6263,
+      "step": 6612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8044357935825248,
+      "learning_rate": 5.112637867792544e-06,
+      "loss": 0.7768,
+      "step": 6613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5207172511477907,
+      "learning_rate": 5.109763946054163e-06,
+      "loss": 0.7245,
+      "step": 6614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6838757695281426,
+      "learning_rate": 5.106890555074426e-06,
+      "loss": 0.6495,
+      "step": 6615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5251784247030207,
+      "learning_rate": 5.1040176951652046e-06,
+      "loss": 0.668,
+      "step": 6616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.965638309649969,
+      "learning_rate": 5.101145366638294e-06,
+      "loss": 0.6732,
+      "step": 6617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6207157830214798,
+      "learning_rate": 5.098273569805443e-06,
+      "loss": 0.6844,
+      "step": 6618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4730282022868064,
+      "learning_rate": 5.095402304978346e-06,
+      "loss": 0.6946,
+      "step": 6619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5416613162094852,
+      "learning_rate": 5.092531572468627e-06,
+      "loss": 0.6273,
+      "step": 6620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.591078859736904,
+      "learning_rate": 5.089661372587859e-06,
+      "loss": 0.7806,
+      "step": 6621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6373143476009928,
+      "learning_rate": 5.086791705647566e-06,
+      "loss": 0.7163,
+      "step": 6622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5145191628770813,
+      "learning_rate": 5.083922571959194e-06,
+      "loss": 0.5919,
+      "step": 6623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4928294951142096,
+      "learning_rate": 5.081053971834153e-06,
+      "loss": 0.6832,
+      "step": 6624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.551921705364535,
+      "learning_rate": 5.078185905583774e-06,
+      "loss": 0.6539,
+      "step": 6625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6046975733459894,
+      "learning_rate": 5.075318373519349e-06,
+      "loss": 0.7742,
+      "step": 6626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6771471734669585,
+      "learning_rate": 5.072451375952104e-06,
+      "loss": 0.7703,
+      "step": 6627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7927702142712978,
+      "learning_rate": 5.069584913193199e-06,
+      "loss": 0.812,
+      "step": 6628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6327378040139973,
+      "learning_rate": 5.066718985553749e-06,
+      "loss": 0.6578,
+      "step": 6629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5301086870655425,
+      "learning_rate": 5.06385359334481e-06,
+      "loss": 0.6644,
+      "step": 6630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.553619736705293,
+      "learning_rate": 5.060988736877366e-06,
+      "loss": 0.727,
+      "step": 6631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6295412264746585,
+      "learning_rate": 5.0581244164623576e-06,
+      "loss": 0.687,
+      "step": 6632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4718060176301735,
+      "learning_rate": 5.055260632410665e-06,
+      "loss": 0.6943,
+      "step": 6633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5675917131961066,
+      "learning_rate": 5.0523973850331e-06,
+      "loss": 0.7724,
+      "step": 6634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.678519567246987,
+      "learning_rate": 5.049534674640426e-06,
+      "loss": 0.6734,
+      "step": 6635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6314176517551688,
+      "learning_rate": 5.0466725015433526e-06,
+      "loss": 0.6701,
+      "step": 6636
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6684931385240929,
+      "learning_rate": 5.043810866052512e-06,
+      "loss": 0.7685,
+      "step": 6637
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.559866500388052,
+      "learning_rate": 5.040949768478496e-06,
+      "loss": 0.7064,
+      "step": 6638
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6826584219645397,
+      "learning_rate": 5.038089209131837e-06,
+      "loss": 0.7978,
+      "step": 6639
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8855229766745862,
+      "learning_rate": 5.035229188322994e-06,
+      "loss": 0.7099,
+      "step": 6640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6390837167916354,
+      "learning_rate": 5.032369706362381e-06,
+      "loss": 0.7395,
+      "step": 6641
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.501803420191579,
+      "learning_rate": 5.029510763560357e-06,
+      "loss": 0.7012,
+      "step": 6642
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4287509327965948,
+      "learning_rate": 5.026652360227205e-06,
+      "loss": 0.6711,
+      "step": 6643
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6337553722301748,
+      "learning_rate": 5.023794496673169e-06,
+      "loss": 0.597,
+      "step": 6644
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6900452701037583,
+      "learning_rate": 5.020937173208418e-06,
+      "loss": 0.7713,
+      "step": 6645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.630566533375451,
+      "learning_rate": 5.018080390143072e-06,
+      "loss": 0.7175,
+      "step": 6646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6482122448531207,
+      "learning_rate": 5.0152241477871955e-06,
+      "loss": 0.665,
+      "step": 6647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5223059886944974,
+      "learning_rate": 5.0123684464507805e-06,
+      "loss": 0.6861,
+      "step": 6648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5625226235889387,
+      "learning_rate": 5.009513286443773e-06,
+      "loss": 0.7124,
+      "step": 6649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6641033534342602,
+      "learning_rate": 5.00665866807606e-06,
+      "loss": 0.6798,
+      "step": 6650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.814997332535629,
+      "learning_rate": 5.003804591657457e-06,
+      "loss": 0.6784,
+      "step": 6651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7418241486270254,
+      "learning_rate": 5.0009510574977385e-06,
+      "loss": 0.684,
+      "step": 6652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8990670470019306,
+      "learning_rate": 4.998098065906601e-06,
+      "loss": 0.7724,
+      "step": 6653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5522110614566593,
+      "learning_rate": 4.9952456171936985e-06,
+      "loss": 0.6414,
+      "step": 6654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6383848530820382,
+      "learning_rate": 4.992393711668622e-06,
+      "loss": 0.8211,
+      "step": 6655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7466103343291657,
+      "learning_rate": 4.989542349640894e-06,
+      "loss": 0.7061,
+      "step": 6656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.569027050350729,
+      "learning_rate": 4.986691531419989e-06,
+      "loss": 0.5821,
+      "step": 6657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6108341079235893,
+      "learning_rate": 4.983841257315324e-06,
+      "loss": 0.7087,
+      "step": 6658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.786948163093711,
+      "learning_rate": 4.980991527636242e-06,
+      "loss": 0.6429,
+      "step": 6659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.698101564669535,
+      "learning_rate": 4.978142342692045e-06,
+      "loss": 0.7126,
+      "step": 6660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6742980171896125,
+      "learning_rate": 4.975293702791959e-06,
+      "loss": 0.7271,
+      "step": 6661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8063655791321431,
+      "learning_rate": 4.9724456082451655e-06,
+      "loss": 0.6859,
+      "step": 6662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.692416378690283,
+      "learning_rate": 4.9695980593607815e-06,
+      "loss": 0.7298,
+      "step": 6663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.610922610672971,
+      "learning_rate": 4.9667510564478584e-06,
+      "loss": 0.7034,
+      "step": 6664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6859325226267035,
+      "learning_rate": 4.9639045998154e-06,
+      "loss": 0.7184,
+      "step": 6665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5747449742645898,
+      "learning_rate": 4.961058689772339e-06,
+      "loss": 0.6004,
+      "step": 6666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6240207523487893,
+      "learning_rate": 4.958213326627555e-06,
+      "loss": 0.6998,
+      "step": 6667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5305125840708707,
+      "learning_rate": 4.9553685106898765e-06,
+      "loss": 0.6738,
+      "step": 6668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4862581273327402,
+      "learning_rate": 4.9525242422680504e-06,
+      "loss": 0.6822,
+      "step": 6669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6232042430434463,
+      "learning_rate": 4.9496805216707885e-06,
+      "loss": 0.682,
+      "step": 6670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.579863545696646,
+      "learning_rate": 4.946837349206726e-06,
+      "loss": 0.6512,
+      "step": 6671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6551829456643596,
+      "learning_rate": 4.943994725184445e-06,
+      "loss": 0.759,
+      "step": 6672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8637227571573647,
+      "learning_rate": 4.941152649912473e-06,
+      "loss": 0.7387,
+      "step": 6673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6064385826752752,
+      "learning_rate": 4.938311123699265e-06,
+      "loss": 0.6942,
+      "step": 6674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.514599431310791,
+      "learning_rate": 4.935470146853229e-06,
+      "loss": 0.6767,
+      "step": 6675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6687265203998711,
+      "learning_rate": 4.9326297196827136e-06,
+      "loss": 0.6838,
+      "step": 6676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3655127070699202,
+      "learning_rate": 4.929789842495992e-06,
+      "loss": 0.6231,
+      "step": 6677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.674511965233358,
+      "learning_rate": 4.9269505156012974e-06,
+      "loss": 0.7355,
+      "step": 6678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6532554229523138,
+      "learning_rate": 4.924111739306788e-06,
+      "loss": 0.732,
+      "step": 6679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6648534043225724,
+      "learning_rate": 4.92127351392057e-06,
+      "loss": 0.7031,
+      "step": 6680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6532155219495979,
+      "learning_rate": 4.918435839750695e-06,
+      "loss": 0.6672,
+      "step": 6681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5060944261708848,
+      "learning_rate": 4.9155987171051374e-06,
+      "loss": 0.5954,
+      "step": 6682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4914021316838773,
+      "learning_rate": 4.912762146291833e-06,
+      "loss": 0.7485,
+      "step": 6683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.708142153364821,
+      "learning_rate": 4.909926127618638e-06,
+      "loss": 0.8124,
+      "step": 6684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6513017550261346,
+      "learning_rate": 4.907090661393362e-06,
+      "loss": 0.6309,
+      "step": 6685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7913658290602579,
+      "learning_rate": 4.904255747923754e-06,
+      "loss": 0.7755,
+      "step": 6686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6348298452794083,
+      "learning_rate": 4.901421387517492e-06,
+      "loss": 0.6811,
+      "step": 6687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.582267175662689,
+      "learning_rate": 4.89858758048221e-06,
+      "loss": 0.6593,
+      "step": 6688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5243067426073116,
+      "learning_rate": 4.895754327125465e-06,
+      "loss": 0.7288,
+      "step": 6689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7220747486755543,
+      "learning_rate": 4.8929216277547665e-06,
+      "loss": 0.6891,
+      "step": 6690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6523749858452668,
+      "learning_rate": 4.8900894826775626e-06,
+      "loss": 0.6876,
+      "step": 6691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6800362457047207,
+      "learning_rate": 4.887257892201231e-06,
+      "loss": 0.7344,
+      "step": 6692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6366830126364655,
+      "learning_rate": 4.884426856633102e-06,
+      "loss": 0.7273,
+      "step": 6693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.625082643217557,
+      "learning_rate": 4.881596376280443e-06,
+      "loss": 0.7569,
+      "step": 6694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.601983835016552,
+      "learning_rate": 4.878766451450451e-06,
+      "loss": 0.6197,
+      "step": 6695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.631133494077483,
+      "learning_rate": 4.875937082450276e-06,
+      "loss": 0.7626,
+      "step": 6696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8279187393898182,
+      "learning_rate": 4.873108269586997e-06,
+      "loss": 0.7777,
+      "step": 6697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.499195022793931,
+      "learning_rate": 4.870280013167639e-06,
+      "loss": 0.6354,
+      "step": 6698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5436191227668006,
+      "learning_rate": 4.867452313499169e-06,
+      "loss": 0.7183,
+      "step": 6699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9507237044500054,
+      "learning_rate": 4.864625170888483e-06,
+      "loss": 0.7534,
+      "step": 6700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5253344116402054,
+      "learning_rate": 4.86179858564243e-06,
+      "loss": 0.6205,
+      "step": 6701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5286591265294258,
+      "learning_rate": 4.858972558067784e-06,
+      "loss": 0.5347,
+      "step": 6702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.673034454696896,
+      "learning_rate": 4.856147088471271e-06,
+      "loss": 0.6869,
+      "step": 6703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7823550632594058,
+      "learning_rate": 4.853322177159554e-06,
+      "loss": 0.7187,
+      "step": 6704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7547562334998144,
+      "learning_rate": 4.850497824439226e-06,
+      "loss": 0.7486,
+      "step": 6705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5239934070506733,
+      "learning_rate": 4.847674030616833e-06,
+      "loss": 0.678,
+      "step": 6706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6717769248144259,
+      "learning_rate": 4.844850795998848e-06,
+      "loss": 0.8623,
+      "step": 6707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5968692392461008,
+      "learning_rate": 4.842028120891691e-06,
+      "loss": 0.6768,
+      "step": 6708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.619490677510862,
+      "learning_rate": 4.839206005601724e-06,
+      "loss": 0.6379,
+      "step": 6709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.419424593862716,
+      "learning_rate": 4.8363844504352344e-06,
+      "loss": 0.6016,
+      "step": 6710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.558674176987791,
+      "learning_rate": 4.8335634556984675e-06,
+      "loss": 0.672,
+      "step": 6711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.719086987642181,
+      "learning_rate": 4.8307430216975895e-06,
+      "loss": 0.7243,
+      "step": 6712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6798971927393056,
+      "learning_rate": 4.827923148738718e-06,
+      "loss": 0.7477,
+      "step": 6713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.715498035259579,
+      "learning_rate": 4.825103837127911e-06,
+      "loss": 0.7564,
+      "step": 6714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7866680476785195,
+      "learning_rate": 4.8222850871711525e-06,
+      "loss": 0.7131,
+      "step": 6715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6589252405439832,
+      "learning_rate": 4.819466899174377e-06,
+      "loss": 0.6056,
+      "step": 6716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7596553684310678,
+      "learning_rate": 4.816649273443457e-06,
+      "loss": 0.6624,
+      "step": 6717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.533186019876807,
+      "learning_rate": 4.813832210284198e-06,
+      "loss": 0.6477,
+      "step": 6718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.782257739414974,
+      "learning_rate": 4.8110157100023545e-06,
+      "loss": 0.6781,
+      "step": 6719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5627791612424267,
+      "learning_rate": 4.808199772903604e-06,
+      "loss": 0.6295,
+      "step": 6720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.538510864281181,
+      "learning_rate": 4.805384399293578e-06,
+      "loss": 0.6813,
+      "step": 6721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6580403085905862,
+      "learning_rate": 4.802569589477844e-06,
+      "loss": 0.6827,
+      "step": 6722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7106126858256103,
+      "learning_rate": 4.799755343761899e-06,
+      "loss": 0.7553,
+      "step": 6723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9278062978061696,
+      "learning_rate": 4.796941662451192e-06,
+      "loss": 0.7088,
+      "step": 6724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6350428421712366,
+      "learning_rate": 4.794128545851098e-06,
+      "loss": 0.63,
+      "step": 6725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6086575592589927,
+      "learning_rate": 4.791315994266939e-06,
+      "loss": 0.6599,
+      "step": 6726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4378733825249688,
+      "learning_rate": 4.788504008003978e-06,
+      "loss": 0.6634,
+      "step": 6727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6964357179844956,
+      "learning_rate": 4.785692587367404e-06,
+      "loss": 0.7541,
+      "step": 6728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5927379582068117,
+      "learning_rate": 4.782881732662361e-06,
+      "loss": 0.6434,
+      "step": 6729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5532184256900274,
+      "learning_rate": 4.780071444193917e-06,
+      "loss": 0.5576,
+      "step": 6730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4686464144373634,
+      "learning_rate": 4.7772617222670846e-06,
+      "loss": 0.8035,
+      "step": 6731
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7957799314682596,
+      "learning_rate": 4.7744525671868235e-06,
+      "loss": 0.6975,
+      "step": 6732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8435799373752362,
+      "learning_rate": 4.771643979258013e-06,
+      "loss": 0.7369,
+      "step": 6733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.501883749908581,
+      "learning_rate": 4.768835958785487e-06,
+      "loss": 0.6216,
+      "step": 6734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3488585383477605,
+      "learning_rate": 4.766028506074014e-06,
+      "loss": 0.6155,
+      "step": 6735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0300629350024613,
+      "learning_rate": 4.763221621428294e-06,
+      "loss": 0.7919,
+      "step": 6736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6568561202507486,
+      "learning_rate": 4.760415305152973e-06,
+      "loss": 0.7961,
+      "step": 6737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5372026401250878,
+      "learning_rate": 4.757609557552635e-06,
+      "loss": 0.6851,
+      "step": 6738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.675958938449828,
+      "learning_rate": 4.754804378931794e-06,
+      "loss": 0.7546,
+      "step": 6739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5616244437905444,
+      "learning_rate": 4.7519997695949135e-06,
+      "loss": 0.6774,
+      "step": 6740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6035797884092577,
+      "learning_rate": 4.749195729846392e-06,
+      "loss": 0.7169,
+      "step": 6741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.589931570461331,
+      "learning_rate": 4.746392259990557e-06,
+      "loss": 0.6673,
+      "step": 6742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4731801910893658,
+      "learning_rate": 4.743589360331683e-06,
+      "loss": 0.5595,
+      "step": 6743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.599173395270854,
+      "learning_rate": 4.740787031173989e-06,
+      "loss": 0.7061,
+      "step": 6744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.634886642584961,
+      "learning_rate": 4.737985272821613e-06,
+      "loss": 0.6353,
+      "step": 6745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.457103767397003,
+      "learning_rate": 4.735184085578647e-06,
+      "loss": 0.6493,
+      "step": 6746
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6446556562271464,
+      "learning_rate": 4.732383469749119e-06,
+      "loss": 0.7001,
+      "step": 6747
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.505602973831173,
+      "learning_rate": 4.7295834256369865e-06,
+      "loss": 0.7058,
+      "step": 6748
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7229690110012572,
+      "learning_rate": 4.726783953546156e-06,
+      "loss": 0.7497,
+      "step": 6749
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7983347955264688,
+      "learning_rate": 4.72398505378046e-06,
+      "loss": 0.7001,
+      "step": 6750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5291712364707872,
+      "learning_rate": 4.721186726643679e-06,
+      "loss": 0.7008,
+      "step": 6751
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5772590436598817,
+      "learning_rate": 4.718388972439531e-06,
+      "loss": 0.6702,
+      "step": 6752
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6705254133139626,
+      "learning_rate": 4.715591791471661e-06,
+      "loss": 0.7293,
+      "step": 6753
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4528902800979537,
+      "learning_rate": 4.712795184043664e-06,
+      "loss": 0.7014,
+      "step": 6754
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8495839131221563,
+      "learning_rate": 4.709999150459069e-06,
+      "loss": 0.7272,
+      "step": 6755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8308693369169993,
+      "learning_rate": 4.707203691021338e-06,
+      "loss": 0.8219,
+      "step": 6756
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8251269818080302,
+      "learning_rate": 4.704408806033876e-06,
+      "loss": 0.7729,
+      "step": 6757
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5691962502332488,
+      "learning_rate": 4.7016144958000275e-06,
+      "loss": 0.6621,
+      "step": 6758
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5036672090092813,
+      "learning_rate": 4.698820760623064e-06,
+      "loss": 0.759,
+      "step": 6759
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6126843045237915,
+      "learning_rate": 4.69602760080621e-06,
+      "loss": 0.7757,
+      "step": 6760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5749756295000583,
+      "learning_rate": 4.693235016652611e-06,
+      "loss": 0.7203,
+      "step": 6761
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5723201831816331,
+      "learning_rate": 4.690443008465363e-06,
+      "loss": 0.7017,
+      "step": 6762
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6678943205347223,
+      "learning_rate": 4.687651576547497e-06,
+      "loss": 0.7812,
+      "step": 6763
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6563925187150188,
+      "learning_rate": 4.684860721201973e-06,
+      "loss": 0.737,
+      "step": 6764
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6299027094882985,
+      "learning_rate": 4.682070442731703e-06,
+      "loss": 0.6962,
+      "step": 6765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6136354271562663,
+      "learning_rate": 4.679280741439519e-06,
+      "loss": 0.7279,
+      "step": 6766
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5619144216443062,
+      "learning_rate": 4.676491617628204e-06,
+      "loss": 0.7066,
+      "step": 6767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.691344837835973,
+      "learning_rate": 4.673703071600477e-06,
+      "loss": 0.7679,
+      "step": 6768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7127039489976044,
+      "learning_rate": 4.670915103658984e-06,
+      "loss": 0.7076,
+      "step": 6769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5145736321312864,
+      "learning_rate": 4.6681277141063185e-06,
+      "loss": 0.7129,
+      "step": 6770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6572377542918708,
+      "learning_rate": 4.665340903245013e-06,
+      "loss": 0.7299,
+      "step": 6771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7087069822360468,
+      "learning_rate": 4.662554671377523e-06,
+      "loss": 0.6861,
+      "step": 6772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5534019045426446,
+      "learning_rate": 4.659769018806259e-06,
+      "loss": 0.6448,
+      "step": 6773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.697895456294768,
+      "learning_rate": 4.656983945833553e-06,
+      "loss": 0.7074,
+      "step": 6774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5580354877752964,
+      "learning_rate": 4.654199452761684e-06,
+      "loss": 0.6663,
+      "step": 6775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7294341968255658,
+      "learning_rate": 4.65141553989287e-06,
+      "loss": 0.6761,
+      "step": 6776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4565405990269753,
+      "learning_rate": 4.648632207529251e-06,
+      "loss": 0.6189,
+      "step": 6777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.567351511841771,
+      "learning_rate": 4.645849455972925e-06,
+      "loss": 0.7025,
+      "step": 6778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6730705756743636,
+      "learning_rate": 4.643067285525907e-06,
+      "loss": 0.7288,
+      "step": 6779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4232719505119826,
+      "learning_rate": 4.640285696490162e-06,
+      "loss": 0.7396,
+      "step": 6780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7484179890091003,
+      "learning_rate": 4.637504689167593e-06,
+      "loss": 0.772,
+      "step": 6781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4096556326775183,
+      "learning_rate": 4.634724263860024e-06,
+      "loss": 0.6992,
+      "step": 6782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.431220240140526,
+      "learning_rate": 4.631944420869237e-06,
+      "loss": 0.6121,
+      "step": 6783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5569615212600543,
+      "learning_rate": 4.629165160496933e-06,
+      "loss": 0.7122,
+      "step": 6784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.61376524661055,
+      "learning_rate": 4.62638648304476e-06,
+      "loss": 0.7074,
+      "step": 6785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.758923600513649,
+      "learning_rate": 4.623608388814305e-06,
+      "loss": 0.7616,
+      "step": 6786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8161912002507001,
+      "learning_rate": 4.620830878107076e-06,
+      "loss": 0.6772,
+      "step": 6787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5930783730088638,
+      "learning_rate": 4.618053951224537e-06,
+      "loss": 0.5818,
+      "step": 6788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6296840520642266,
+      "learning_rate": 4.6152776084680795e-06,
+      "loss": 0.6261,
+      "step": 6789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.637986537306099,
+      "learning_rate": 4.612501850139027e-06,
+      "loss": 0.6713,
+      "step": 6790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7392169691741854,
+      "learning_rate": 4.609726676538652e-06,
+      "loss": 0.669,
+      "step": 6791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6813433964446511,
+      "learning_rate": 4.6069520879681464e-06,
+      "loss": 0.6574,
+      "step": 6792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.566228995523543,
+      "learning_rate": 4.604178084728655e-06,
+      "loss": 0.7067,
+      "step": 6793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6410680316688766,
+      "learning_rate": 4.601404667121254e-06,
+      "loss": 0.7132,
+      "step": 6794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6200882326123738,
+      "learning_rate": 4.5986318354469495e-06,
+      "loss": 0.6418,
+      "step": 6795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.846150123331626,
+      "learning_rate": 4.595859590006694e-06,
+      "loss": 0.8386,
+      "step": 6796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4435592575044285,
+      "learning_rate": 4.593087931101366e-06,
+      "loss": 0.6287,
+      "step": 6797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5756239357254955,
+      "learning_rate": 4.590316859031787e-06,
+      "loss": 0.606,
+      "step": 6798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5562145930337141,
+      "learning_rate": 4.5875463740987196e-06,
+      "loss": 0.6404,
+      "step": 6799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7451565402362117,
+      "learning_rate": 4.584776476602848e-06,
+      "loss": 0.7641,
+      "step": 6800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7561251878769575,
+      "learning_rate": 4.582007166844808e-06,
+      "loss": 0.7515,
+      "step": 6801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6443111262582564,
+      "learning_rate": 4.579238445125159e-06,
+      "loss": 0.7593,
+      "step": 6802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8102861573409768,
+      "learning_rate": 4.576470311744406e-06,
+      "loss": 0.7399,
+      "step": 6803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6888475048886598,
+      "learning_rate": 4.573702767002991e-06,
+      "loss": 0.6528,
+      "step": 6804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6537094560055399,
+      "learning_rate": 4.570935811201277e-06,
+      "loss": 0.743,
+      "step": 6805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.609274630536314,
+      "learning_rate": 4.5681694446395805e-06,
+      "loss": 0.7941,
+      "step": 6806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6042699551754307,
+      "learning_rate": 4.5654036676181505e-06,
+      "loss": 0.7543,
+      "step": 6807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6824929983387507,
+      "learning_rate": 4.5626384804371605e-06,
+      "loss": 0.8842,
+      "step": 6808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.48386149237531,
+      "learning_rate": 4.559873883396737e-06,
+      "loss": 0.5557,
+      "step": 6809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.541725899034693,
+      "learning_rate": 4.557109876796926e-06,
+      "loss": 0.6702,
+      "step": 6810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.71511462621031,
+      "learning_rate": 4.554346460937719e-06,
+      "loss": 0.6853,
+      "step": 6811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5697263264236596,
+      "learning_rate": 4.551583636119048e-06,
+      "loss": 0.6495,
+      "step": 6812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.593495035401068,
+      "learning_rate": 4.548821402640765e-06,
+      "loss": 0.8006,
+      "step": 6813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5161862699596542,
+      "learning_rate": 4.5460597608026765e-06,
+      "loss": 0.7516,
+      "step": 6814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7836780242640413,
+      "learning_rate": 4.543298710904507e-06,
+      "loss": 0.6723,
+      "step": 6815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6477562008663253,
+      "learning_rate": 4.54053825324593e-06,
+      "loss": 0.6872,
+      "step": 6816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6047023712842892,
+      "learning_rate": 4.537778388126552e-06,
+      "loss": 0.6548,
+      "step": 6817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.59282147631226,
+      "learning_rate": 4.535019115845905e-06,
+      "loss": 0.6552,
+      "step": 6818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5241484978282966,
+      "learning_rate": 4.532260436703476e-06,
+      "loss": 0.6947,
+      "step": 6819
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.629690488840868,
+      "learning_rate": 4.529502350998667e-06,
+      "loss": 0.6956,
+      "step": 6820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7973943291335965,
+      "learning_rate": 4.526744859030828e-06,
+      "loss": 0.7769,
+      "step": 6821
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8716176708220122,
+      "learning_rate": 4.523987961099247e-06,
+      "loss": 0.8152,
+      "step": 6822
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7628031001157096,
+      "learning_rate": 4.5212316575031325e-06,
+      "loss": 0.7304,
+      "step": 6823
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6365695307885848,
+      "learning_rate": 4.518475948541644e-06,
+      "loss": 0.626,
+      "step": 6824
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8518163270282286,
+      "learning_rate": 4.5157208345138735e-06,
+      "loss": 0.7279,
+      "step": 6825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.503081942774786,
+      "learning_rate": 4.5129663157188385e-06,
+      "loss": 0.6428,
+      "step": 6826
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6458945015573683,
+      "learning_rate": 4.510212392455506e-06,
+      "loss": 0.6701,
+      "step": 6827
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.528661732099273,
+      "learning_rate": 4.507459065022764e-06,
+      "loss": 0.6397,
+      "step": 6828
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7525925171036087,
+      "learning_rate": 4.504706333719446e-06,
+      "loss": 0.7344,
+      "step": 6829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6132996914626434,
+      "learning_rate": 4.501954198844323e-06,
+      "loss": 0.6716,
+      "step": 6830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.728012838361854,
+      "learning_rate": 4.499202660696088e-06,
+      "loss": 0.6334,
+      "step": 6831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6940877753151904,
+      "learning_rate": 4.4964517195733845e-06,
+      "loss": 0.7678,
+      "step": 6832
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4309951646007812,
+      "learning_rate": 4.493701375774778e-06,
+      "loss": 0.6187,
+      "step": 6833
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6994002275015478,
+      "learning_rate": 4.490951629598779e-06,
+      "loss": 0.6717,
+      "step": 6834
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5225830335907207,
+      "learning_rate": 4.488202481343831e-06,
+      "loss": 0.6211,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7419741594304876,
+      "learning_rate": 4.485453931308306e-06,
+      "loss": 0.7343,
+      "step": 6836
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7353467457564924,
+      "learning_rate": 4.482705979790522e-06,
+      "loss": 0.623,
+      "step": 6837
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6379948590288491,
+      "learning_rate": 4.479958627088719e-06,
+      "loss": 0.626,
+      "step": 6838
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5886343067209587,
+      "learning_rate": 4.477211873501085e-06,
+      "loss": 0.6164,
+      "step": 6839
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6088601476355608,
+      "learning_rate": 4.474465719325738e-06,
+      "loss": 0.5527,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4754554021635713,
+      "learning_rate": 4.471720164860723e-06,
+      "loss": 0.6268,
+      "step": 6841
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6109931340884776,
+      "learning_rate": 4.468975210404032e-06,
+      "loss": 0.7196,
+      "step": 6842
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4207320523598963,
+      "learning_rate": 4.4662308562535905e-06,
+      "loss": 0.6549,
+      "step": 6843
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5971967413292498,
+      "learning_rate": 4.463487102707247e-06,
+      "loss": 0.7004,
+      "step": 6844
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5846312944544336,
+      "learning_rate": 4.460743950062797e-06,
+      "loss": 0.6683,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5170102406603916,
+      "learning_rate": 4.458001398617971e-06,
+      "loss": 0.6649,
+      "step": 6846
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6523674664305048,
+      "learning_rate": 4.455259448670423e-06,
+      "loss": 0.7747,
+      "step": 6847
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5808458259236084,
+      "learning_rate": 4.452518100517751e-06,
+      "loss": 0.7314,
+      "step": 6848
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5507415500961412,
+      "learning_rate": 4.449777354457489e-06,
+      "loss": 0.6792,
+      "step": 6849
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6687667637042354,
+      "learning_rate": 4.447037210787098e-06,
+      "loss": 0.6677,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6465995920302823,
+      "learning_rate": 4.444297669803981e-06,
+      "loss": 0.688,
+      "step": 6851
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3837375906710079,
+      "learning_rate": 4.441558731805467e-06,
+      "loss": 0.6833,
+      "step": 6852
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6111410414667975,
+      "learning_rate": 4.438820397088829e-06,
+      "loss": 0.7622,
+      "step": 6853
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7558194753522183,
+      "learning_rate": 4.436082665951272e-06,
+      "loss": 0.7165,
+      "step": 6854
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7187189582051745,
+      "learning_rate": 4.43334553868993e-06,
+      "loss": 0.7752,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5985573828842572,
+      "learning_rate": 4.430609015601875e-06,
+      "loss": 0.7959,
+      "step": 6856
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6240805305240187,
+      "learning_rate": 4.427873096984122e-06,
+      "loss": 0.7092,
+      "step": 6857
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6149474725862267,
+      "learning_rate": 4.4251377831336005e-06,
+      "loss": 0.711,
+      "step": 6858
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5581486292014552,
+      "learning_rate": 4.422403074347192e-06,
+      "loss": 0.7262,
+      "step": 6859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6611056685471053,
+      "learning_rate": 4.41966897092171e-06,
+      "loss": 0.7394,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6201934588493567,
+      "learning_rate": 4.416935473153891e-06,
+      "loss": 0.7084,
+      "step": 6861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.698794288828691,
+      "learning_rate": 4.414202581340416e-06,
+      "loss": 0.7496,
+      "step": 6862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5843943183950784,
+      "learning_rate": 4.411470295777904e-06,
+      "loss": 0.7264,
+      "step": 6863
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7047690072131172,
+      "learning_rate": 4.4087386167628905e-06,
+      "loss": 0.6368,
+      "step": 6864
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5051173591130502,
+      "learning_rate": 4.4060075445918635e-06,
+      "loss": 0.672,
+      "step": 6865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6290516020944361,
+      "learning_rate": 4.4032770795612426e-06,
+      "loss": 0.6965,
+      "step": 6866
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4918612948316938,
+      "learning_rate": 4.400547221967366e-06,
+      "loss": 0.7075,
+      "step": 6867
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5594186028274144,
+      "learning_rate": 4.397817972106527e-06,
+      "loss": 0.6735,
+      "step": 6868
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4450516338889967,
+      "learning_rate": 4.395089330274936e-06,
+      "loss": 0.5918,
+      "step": 6869
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5175043132909714,
+      "learning_rate": 4.392361296768747e-06,
+      "loss": 0.7221,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6293215499348013,
+      "learning_rate": 4.389633871884048e-06,
+      "loss": 0.7563,
+      "step": 6871
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4893946119439887,
+      "learning_rate": 4.386907055916854e-06,
+      "loss": 0.5862,
+      "step": 6872
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5866286954728583,
+      "learning_rate": 4.384180849163122e-06,
+      "loss": 0.6307,
+      "step": 6873
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.533145190738263,
+      "learning_rate": 4.381455251918735e-06,
+      "loss": 0.6756,
+      "step": 6874
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6868162377836897,
+      "learning_rate": 4.378730264479516e-06,
+      "loss": 0.6691,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6819700258658672,
+      "learning_rate": 4.376005887141223e-06,
+      "loss": 0.6826,
+      "step": 6876
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7399392860899976,
+      "learning_rate": 4.373282120199538e-06,
+      "loss": 0.6818,
+      "step": 6877
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7575382587651571,
+      "learning_rate": 4.370558963950091e-06,
+      "loss": 0.7339,
+      "step": 6878
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.790098215420934,
+      "learning_rate": 4.36783641868843e-06,
+      "loss": 0.7084,
+      "step": 6879
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7916971074702635,
+      "learning_rate": 4.365114484710048e-06,
+      "loss": 0.7472,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.634292686325334,
+      "learning_rate": 4.3623931623103735e-06,
+      "loss": 0.611,
+      "step": 6881
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5793655961743505,
+      "learning_rate": 4.359672451784754e-06,
+      "loss": 0.624,
+      "step": 6882
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.750260127143254,
+      "learning_rate": 4.356952353428486e-06,
+      "loss": 0.7962,
+      "step": 6883
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7649362401639273,
+      "learning_rate": 4.354232867536797e-06,
+      "loss": 0.7676,
+      "step": 6884
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.752447193640601,
+      "learning_rate": 4.351513994404835e-06,
+      "loss": 0.7463,
+      "step": 6885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5594556890072002,
+      "learning_rate": 4.348795734327701e-06,
+      "loss": 0.6703,
+      "step": 6886
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.648662065675375,
+      "learning_rate": 4.346078087600411e-06,
+      "loss": 0.7742,
+      "step": 6887
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.699704656113283,
+      "learning_rate": 4.343361054517927e-06,
+      "loss": 0.7956,
+      "step": 6888
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6594430543250784,
+      "learning_rate": 4.340644635375146e-06,
+      "loss": 0.7169,
+      "step": 6889
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7104654398985026,
+      "learning_rate": 4.337928830466882e-06,
+      "loss": 0.7365,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.755419660001335,
+      "learning_rate": 4.335213640087902e-06,
+      "loss": 0.7746,
+      "step": 6891
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6038165272888565,
+      "learning_rate": 4.332499064532891e-06,
+      "loss": 0.6323,
+      "step": 6892
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7456593937639815,
+      "learning_rate": 4.329785104096477e-06,
+      "loss": 0.7454,
+      "step": 6893
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6068745247970297,
+      "learning_rate": 4.327071759073221e-06,
+      "loss": 0.6262,
+      "step": 6894
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6818545930186335,
+      "learning_rate": 4.324359029757607e-06,
+      "loss": 0.6403,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4850855188004741,
+      "learning_rate": 4.321646916444068e-06,
+      "loss": 0.6217,
+      "step": 6896
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7718084166975085,
+      "learning_rate": 4.318935419426951e-06,
+      "loss": 0.7067,
+      "step": 6897
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5457835040849344,
+      "learning_rate": 4.316224539000554e-06,
+      "loss": 0.6807,
+      "step": 6898
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7754546506945497,
+      "learning_rate": 4.313514275459103e-06,
+      "loss": 0.8154,
+      "step": 6899
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6481037379539556,
+      "learning_rate": 4.310804629096746e-06,
+      "loss": 0.7368,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6254104503335625,
+      "learning_rate": 4.308095600207578e-06,
+      "loss": 0.6914,
+      "step": 6901
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6361449332363762,
+      "learning_rate": 4.3053871890856235e-06,
+      "loss": 0.6178,
+      "step": 6902
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4364947519767834,
+      "learning_rate": 4.302679396024834e-06,
+      "loss": 0.6167,
+      "step": 6903
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6631765011083663,
+      "learning_rate": 4.299972221319103e-06,
+      "loss": 0.6483,
+      "step": 6904
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6516281319651802,
+      "learning_rate": 4.2972656652622446e-06,
+      "loss": 0.769,
+      "step": 6905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5056024308236098,
+      "learning_rate": 4.294559728148018e-06,
+      "loss": 0.6928,
+      "step": 6906
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6055309450592068,
+      "learning_rate": 4.291854410270113e-06,
+      "loss": 0.7404,
+      "step": 6907
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.671936816083641,
+      "learning_rate": 4.289149711922143e-06,
+      "loss": 0.746,
+      "step": 6908
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4401133849964285,
+      "learning_rate": 4.2864456333976676e-06,
+      "loss": 0.6935,
+      "step": 6909
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.663931022188339,
+      "learning_rate": 4.2837421749901655e-06,
+      "loss": 0.7025,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6514456944306992,
+      "learning_rate": 4.281039336993058e-06,
+      "loss": 0.7564,
+      "step": 6911
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.590988972850426,
+      "learning_rate": 4.278337119699699e-06,
+      "loss": 0.6574,
+      "step": 6912
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7075607813218283,
+      "learning_rate": 4.275635523403367e-06,
+      "loss": 0.6821,
+      "step": 6913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6835764812133978,
+      "learning_rate": 4.27293454839728e-06,
+      "loss": 0.7864,
+      "step": 6914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6808712795164975,
+      "learning_rate": 4.270234194974586e-06,
+      "loss": 0.7632,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6984519613192404,
+      "learning_rate": 4.267534463428365e-06,
+      "loss": 0.7588,
+      "step": 6916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7880445241677712,
+      "learning_rate": 4.264835354051636e-06,
+      "loss": 0.7459,
+      "step": 6917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.524997395722238,
+      "learning_rate": 4.262136867137339e-06,
+      "loss": 0.7045,
+      "step": 6918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8045010726243293,
+      "learning_rate": 4.259439002978354e-06,
+      "loss": 0.7398,
+      "step": 6919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5151094218387287,
+      "learning_rate": 4.256741761867497e-06,
+      "loss": 0.6549,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6018412694118114,
+      "learning_rate": 4.2540451440975025e-06,
+      "loss": 0.7484,
+      "step": 6921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5854447630426047,
+      "learning_rate": 4.251349149961056e-06,
+      "loss": 0.7279,
+      "step": 6922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4938982874778577,
+      "learning_rate": 4.248653779750756e-06,
+      "loss": 0.6409,
+      "step": 6923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5468644761910109,
+      "learning_rate": 4.245959033759147e-06,
+      "loss": 0.6326,
+      "step": 6924
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.666205435921289,
+      "learning_rate": 4.243264912278706e-06,
+      "loss": 0.804,
+      "step": 6925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6496617351717349,
+      "learning_rate": 4.2405714156018295e-06,
+      "loss": 0.7252,
+      "step": 6926
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5762382787063267,
+      "learning_rate": 4.237878544020861e-06,
+      "loss": 0.6196,
+      "step": 6927
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7349560074224841,
+      "learning_rate": 4.235186297828066e-06,
+      "loss": 0.682,
+      "step": 6928
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6572899000817862,
+      "learning_rate": 4.232494677315645e-06,
+      "loss": 0.6835,
+      "step": 6929
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.70827240724343,
+      "learning_rate": 4.2298036827757375e-06,
+      "loss": 0.7192,
+      "step": 6930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9011840639795596,
+      "learning_rate": 4.2271133145003995e-06,
+      "loss": 0.7168,
+      "step": 6931
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5761513167033099,
+      "learning_rate": 4.224423572781637e-06,
+      "loss": 0.768,
+      "step": 6932
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5980729940148364,
+      "learning_rate": 4.221734457911374e-06,
+      "loss": 0.6542,
+      "step": 6933
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.577183650117114,
+      "learning_rate": 4.219045970181472e-06,
+      "loss": 0.6186,
+      "step": 6934
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5433345907486062,
+      "learning_rate": 4.21635810988373e-06,
+      "loss": 0.6485,
+      "step": 6935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6069760167637266,
+      "learning_rate": 4.213670877309867e-06,
+      "loss": 0.6646,
+      "step": 6936
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7019206675257987,
+      "learning_rate": 4.210984272751541e-06,
+      "loss": 0.6954,
+      "step": 6937
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7657777698386452,
+      "learning_rate": 4.208298296500345e-06,
+      "loss": 0.7059,
+      "step": 6938
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5774180458446942,
+      "learning_rate": 4.2056129488477936e-06,
+      "loss": 0.593,
+      "step": 6939
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6366090495196506,
+      "learning_rate": 4.2029282300853455e-06,
+      "loss": 0.6479,
+      "step": 6940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6983953653821355,
+      "learning_rate": 4.200244140504379e-06,
+      "loss": 0.7052,
+      "step": 6941
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6921720513226657,
+      "learning_rate": 4.197560680396214e-06,
+      "loss": 0.7774,
+      "step": 6942
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5070654563971895,
+      "learning_rate": 4.1948778500521e-06,
+      "loss": 0.6766,
+      "step": 6943
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.518188850551967,
+      "learning_rate": 4.19219564976321e-06,
+      "loss": 0.7744,
+      "step": 6944
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.479936391862102,
+      "learning_rate": 4.189514079820662e-06,
+      "loss": 0.6599,
+      "step": 6945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6671871455146776,
+      "learning_rate": 4.1868331405154905e-06,
+      "loss": 0.6885,
+      "step": 6946
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7031726324212713,
+      "learning_rate": 4.184152832138673e-06,
+      "loss": 0.7379,
+      "step": 6947
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5421456573559247,
+      "learning_rate": 4.181473154981122e-06,
+      "loss": 0.5938,
+      "step": 6948
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7023553958191253,
+      "learning_rate": 4.1787941093336635e-06,
+      "loss": 0.7094,
+      "step": 6949
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4727549625930916,
+      "learning_rate": 4.176115695487071e-06,
+      "loss": 0.6625,
+      "step": 6950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7742791080305533,
+      "learning_rate": 4.173437913732048e-06,
+      "loss": 0.7806,
+      "step": 6951
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6522047579037553,
+      "learning_rate": 4.1707607643592185e-06,
+      "loss": 0.6883,
+      "step": 6952
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5643275729858779,
+      "learning_rate": 4.1680842476591484e-06,
+      "loss": 0.7131,
+      "step": 6953
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6108363196318192,
+      "learning_rate": 4.165408363922337e-06,
+      "loss": 0.7833,
+      "step": 6954
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7994867885909294,
+      "learning_rate": 4.1627331134392e-06,
+      "loss": 0.6775,
+      "step": 6955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7217488003033425,
+      "learning_rate": 4.160058496500103e-06,
+      "loss": 0.6403,
+      "step": 6956
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6425976302194905,
+      "learning_rate": 4.1573845133953275e-06,
+      "loss": 0.7085,
+      "step": 6957
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.497892765596392,
+      "learning_rate": 4.154711164415094e-06,
+      "loss": 0.6671,
+      "step": 6958
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8054907860900105,
+      "learning_rate": 4.152038449849556e-06,
+      "loss": 0.8228,
+      "step": 6959
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.562254776905117,
+      "learning_rate": 4.14936636998879e-06,
+      "loss": 0.6295,
+      "step": 6960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7181351357516357,
+      "learning_rate": 4.14669492512281e-06,
+      "loss": 0.7395,
+      "step": 6961
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6343855996101109,
+      "learning_rate": 4.144024115541565e-06,
+      "loss": 0.7496,
+      "step": 6962
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6993064244870406,
+      "learning_rate": 4.14135394153492e-06,
+      "loss": 0.6616,
+      "step": 6963
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.543490876092734,
+      "learning_rate": 4.138684403392688e-06,
+      "loss": 0.7636,
+      "step": 6964
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5772250311731004,
+      "learning_rate": 4.136015501404604e-06,
+      "loss": 0.6949,
+      "step": 6965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0741991444368217,
+      "learning_rate": 4.133347235860333e-06,
+      "loss": 0.863,
+      "step": 6966
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.647681598984127,
+      "learning_rate": 4.1306796070494755e-06,
+      "loss": 0.7291,
+      "step": 6967
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5520196505035508,
+      "learning_rate": 4.1280126152615644e-06,
+      "loss": 0.6729,
+      "step": 6968
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5718487268044565,
+      "learning_rate": 4.1253462607860515e-06,
+      "loss": 0.6882,
+      "step": 6969
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6006646987325581,
+      "learning_rate": 4.122680543912334e-06,
+      "loss": 0.6832,
+      "step": 6970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5881086125681252,
+      "learning_rate": 4.120015464929735e-06,
+      "loss": 0.6927,
+      "step": 6971
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6140338625443162,
+      "learning_rate": 4.117351024127501e-06,
+      "loss": 0.7615,
+      "step": 6972
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7039718584416361,
+      "learning_rate": 4.11468722179482e-06,
+      "loss": 0.8646,
+      "step": 6973
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5917611835614012,
+      "learning_rate": 4.112024058220808e-06,
+      "loss": 0.6972,
+      "step": 6974
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.740603652542512,
+      "learning_rate": 4.109361533694504e-06,
+      "loss": 0.7257,
+      "step": 6975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7885656752367534,
+      "learning_rate": 4.10669964850489e-06,
+      "loss": 0.7949,
+      "step": 6976
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7294983851635601,
+      "learning_rate": 4.104038402940863e-06,
+      "loss": 0.7238,
+      "step": 6977
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6325619934130196,
+      "learning_rate": 4.101377797291265e-06,
+      "loss": 0.6057,
+      "step": 6978
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5344764459127072,
+      "learning_rate": 4.098717831844867e-06,
+      "loss": 0.6152,
+      "step": 6979
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6435855427908748,
+      "learning_rate": 4.09605850689036e-06,
+      "loss": 0.6682,
+      "step": 6980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9189032336719265,
+      "learning_rate": 4.093399822716377e-06,
+      "loss": 0.6975,
+      "step": 6981
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6852666215395304,
+      "learning_rate": 4.090741779611472e-06,
+      "loss": 0.7259,
+      "step": 6982
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.710324596376418,
+      "learning_rate": 4.088084377864136e-06,
+      "loss": 0.767,
+      "step": 6983
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6380525152065986,
+      "learning_rate": 4.085427617762792e-06,
+      "loss": 0.6968,
+      "step": 6984
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5433413494513588,
+      "learning_rate": 4.082771499595782e-06,
+      "loss": 0.6676,
+      "step": 6985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.779847131311914,
+      "learning_rate": 4.080116023651396e-06,
+      "loss": 0.6974,
+      "step": 6986
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5818748331761248,
+      "learning_rate": 4.077461190217835e-06,
+      "loss": 0.7561,
+      "step": 6987
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7205646792957754,
+      "learning_rate": 4.074806999583243e-06,
+      "loss": 0.7579,
+      "step": 6988
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.503529710542259,
+      "learning_rate": 4.0721534520356955e-06,
+      "loss": 0.6626,
+      "step": 6989
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7433595830714752,
+      "learning_rate": 4.069500547863185e-06,
+      "loss": 0.7432,
+      "step": 6990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5407275190849465,
+      "learning_rate": 4.06684828735365e-06,
+      "loss": 0.6634,
+      "step": 6991
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.879585303039927,
+      "learning_rate": 4.0641966707949506e-06,
+      "loss": 0.7283,
+      "step": 6992
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6665885895090278,
+      "learning_rate": 4.061545698474875e-06,
+      "loss": 0.7186,
+      "step": 6993
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6482473801952147,
+      "learning_rate": 4.0588953706811485e-06,
+      "loss": 0.6727,
+      "step": 6994
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6151987819235434,
+      "learning_rate": 4.056245687701419e-06,
+      "loss": 0.6786,
+      "step": 6995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.723556396094163,
+      "learning_rate": 4.053596649823269e-06,
+      "loss": 0.6509,
+      "step": 6996
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7011356810748508,
+      "learning_rate": 4.050948257334215e-06,
+      "loss": 0.8039,
+      "step": 6997
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6895154305021753,
+      "learning_rate": 4.048300510521692e-06,
+      "loss": 0.6526,
+      "step": 6998
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9452574818201525,
+      "learning_rate": 4.045653409673079e-06,
+      "loss": 0.8085,
+      "step": 6999
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7609192721254838,
+      "learning_rate": 4.043006955075667e-06,
+      "loss": 0.7762,
+      "step": 7000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7793784128833332,
+      "learning_rate": 4.040361147016693e-06,
+      "loss": 0.7615,
+      "step": 7001
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7238975070132547,
+      "learning_rate": 4.0377159857833235e-06,
+      "loss": 0.6991,
+      "step": 7002
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6611134034266273,
+      "learning_rate": 4.035071471662641e-06,
+      "loss": 0.7009,
+      "step": 7003
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7967481323766012,
+      "learning_rate": 4.032427604941671e-06,
+      "loss": 0.6879,
+      "step": 7004
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6103088335675344,
+      "learning_rate": 4.029784385907359e-06,
+      "loss": 0.6998,
+      "step": 7005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4755229277285078,
+      "learning_rate": 4.027141814846588e-06,
+      "loss": 0.6892,
+      "step": 7006
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7663439626959418,
+      "learning_rate": 4.024499892046172e-06,
+      "loss": 0.6752,
+      "step": 7007
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6168490704977603,
+      "learning_rate": 4.021858617792843e-06,
+      "loss": 0.6625,
+      "step": 7008
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7391448458975851,
+      "learning_rate": 4.019217992373277e-06,
+      "loss": 0.7316,
+      "step": 7009
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7057666540441399,
+      "learning_rate": 4.016578016074064e-06,
+      "loss": 0.7041,
+      "step": 7010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6059359526383257,
+      "learning_rate": 4.0139386891817376e-06,
+      "loss": 0.6557,
+      "step": 7011
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.773911345778964,
+      "learning_rate": 4.011300011982759e-06,
+      "loss": 0.7215,
+      "step": 7012
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4984640233051176,
+      "learning_rate": 4.0086619847635075e-06,
+      "loss": 0.6175,
+      "step": 7013
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7100583414353592,
+      "learning_rate": 4.006024607810304e-06,
+      "loss": 0.7669,
+      "step": 7014
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4668525303618518,
+      "learning_rate": 4.003387881409397e-06,
+      "loss": 0.5917,
+      "step": 7015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4201488465349763,
+      "learning_rate": 4.000751805846956e-06,
+      "loss": 0.6535,
+      "step": 7016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7988283699217553,
+      "learning_rate": 3.998116381409091e-06,
+      "loss": 0.7743,
+      "step": 7017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.761206047677621,
+      "learning_rate": 3.995481608381833e-06,
+      "loss": 0.7344,
+      "step": 7018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6422469485706646,
+      "learning_rate": 3.992847487051144e-06,
+      "loss": 0.7581,
+      "step": 7019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.625424783824051,
+      "learning_rate": 3.990214017702923e-06,
+      "loss": 0.707,
+      "step": 7020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.456008490060363,
+      "learning_rate": 3.987581200622984e-06,
+      "loss": 0.6372,
+      "step": 7021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7653640607260257,
+      "learning_rate": 3.984949036097085e-06,
+      "loss": 0.7238,
+      "step": 7022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.70968000846081,
+      "learning_rate": 3.982317524410901e-06,
+      "loss": 0.6212,
+      "step": 7023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7276547862614866,
+      "learning_rate": 3.979686665850043e-06,
+      "loss": 0.7209,
+      "step": 7024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4841760329433793,
+      "learning_rate": 3.977056460700054e-06,
+      "loss": 0.8146,
+      "step": 7025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7281365965194087,
+      "learning_rate": 3.974426909246395e-06,
+      "loss": 0.6069,
+      "step": 7026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7366179797813863,
+      "learning_rate": 3.971798011774468e-06,
+      "loss": 0.681,
+      "step": 7027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7665892061525008,
+      "learning_rate": 3.969169768569593e-06,
+      "loss": 0.6869,
+      "step": 7028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6786218317981734,
+      "learning_rate": 3.966542179917028e-06,
+      "loss": 0.6813,
+      "step": 7029
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4711116283729369,
+      "learning_rate": 3.963915246101961e-06,
+      "loss": 0.5982,
+      "step": 7030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6434577385828204,
+      "learning_rate": 3.9612889674094955e-06,
+      "loss": 0.7025,
+      "step": 7031
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5554711330154485,
+      "learning_rate": 3.958663344124678e-06,
+      "loss": 0.6731,
+      "step": 7032
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6676558574806524,
+      "learning_rate": 3.956038376532482e-06,
+      "loss": 0.6567,
+      "step": 7033
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7124209469173053,
+      "learning_rate": 3.9534140649177995e-06,
+      "loss": 0.7216,
+      "step": 7034
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7075703905836943,
+      "learning_rate": 3.950790409565467e-06,
+      "loss": 0.7102,
+      "step": 7035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5417142310579948,
+      "learning_rate": 3.948167410760232e-06,
+      "loss": 0.664,
+      "step": 7036
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.461830993842134,
+      "learning_rate": 3.9455450687867845e-06,
+      "loss": 0.6277,
+      "step": 7037
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8940398474818356,
+      "learning_rate": 3.942923383929743e-06,
+      "loss": 0.7572,
+      "step": 7038
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9355738539011664,
+      "learning_rate": 3.940302356473642e-06,
+      "loss": 0.7879,
+      "step": 7039
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.687447276435726,
+      "learning_rate": 3.93768198670296e-06,
+      "loss": 0.6419,
+      "step": 7040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5842593327172416,
+      "learning_rate": 3.935062274902092e-06,
+      "loss": 0.7642,
+      "step": 7041
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5795026594057051,
+      "learning_rate": 3.932443221355369e-06,
+      "loss": 0.6418,
+      "step": 7042
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5554013080977451,
+      "learning_rate": 3.929824826347051e-06,
+      "loss": 0.6573,
+      "step": 7043
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5958131006786735,
+      "learning_rate": 3.927207090161318e-06,
+      "loss": 0.648,
+      "step": 7044
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4619791680331131,
+      "learning_rate": 3.9245900130822915e-06,
+      "loss": 0.6172,
+      "step": 7045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.751601619237473,
+      "learning_rate": 3.921973595394005e-06,
+      "loss": 0.711,
+      "step": 7046
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7066907412184105,
+      "learning_rate": 3.919357837380437e-06,
+      "loss": 0.7086,
+      "step": 7047
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.573108845705456,
+      "learning_rate": 3.916742739325488e-06,
+      "loss": 0.7856,
+      "step": 7048
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6505398116251775,
+      "learning_rate": 3.91412830151298e-06,
+      "loss": 0.6768,
+      "step": 7049
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.565935148019075,
+      "learning_rate": 3.9115145242266714e-06,
+      "loss": 0.7227,
+      "step": 7050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7356480172771545,
+      "learning_rate": 3.9089014077502515e-06,
+      "loss": 0.7685,
+      "step": 7051
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6549946772587585,
+      "learning_rate": 3.906288952367328e-06,
+      "loss": 0.7455,
+      "step": 7052
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6271433182157338,
+      "learning_rate": 3.903677158361445e-06,
+      "loss": 0.6842,
+      "step": 7053
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5938354446085299,
+      "learning_rate": 3.901066026016068e-06,
+      "loss": 0.5795,
+      "step": 7054
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.612843481164936,
+      "learning_rate": 3.898455555614598e-06,
+      "loss": 0.7869,
+      "step": 7055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.514332800714335,
+      "learning_rate": 3.895845747440363e-06,
+      "loss": 0.6233,
+      "step": 7056
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7666438214701161,
+      "learning_rate": 3.89323660177661e-06,
+      "loss": 0.7276,
+      "step": 7057
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7673801040896533,
+      "learning_rate": 3.890628118906525e-06,
+      "loss": 0.6719,
+      "step": 7058
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.618196212982791,
+      "learning_rate": 3.888020299113221e-06,
+      "loss": 0.6988,
+      "step": 7059
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6641889014520415,
+      "learning_rate": 3.885413142679729e-06,
+      "loss": 0.6884,
+      "step": 7060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5790124585813679,
+      "learning_rate": 3.882806649889024e-06,
+      "loss": 0.7119,
+      "step": 7061
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5846068340561161,
+      "learning_rate": 3.8802008210239906e-06,
+      "loss": 0.6524,
+      "step": 7062
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.519531275732951,
+      "learning_rate": 3.877595656367454e-06,
+      "loss": 0.638,
+      "step": 7063
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5115523763776504,
+      "learning_rate": 3.8749911562021705e-06,
+      "loss": 0.6822,
+      "step": 7064
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.764434489462425,
+      "learning_rate": 3.872387320810807e-06,
+      "loss": 0.7642,
+      "step": 7065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5793773972876426,
+      "learning_rate": 3.869784150475977e-06,
+      "loss": 0.6887,
+      "step": 7066
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7045250995664494,
+      "learning_rate": 3.867181645480214e-06,
+      "loss": 0.676,
+      "step": 7067
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5399538449783368,
+      "learning_rate": 3.864579806105974e-06,
+      "loss": 0.6566,
+      "step": 7068
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6457773798036648,
+      "learning_rate": 3.861978632635648e-06,
+      "loss": 0.6843,
+      "step": 7069
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5939853478787616,
+      "learning_rate": 3.859378125351556e-06,
+      "loss": 0.7579,
+      "step": 7070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8740643030417445,
+      "learning_rate": 3.856778284535938e-06,
+      "loss": 0.7391,
+      "step": 7071
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6024680692373516,
+      "learning_rate": 3.854179110470968e-06,
+      "loss": 0.7591,
+      "step": 7072
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8009607749236867,
+      "learning_rate": 3.851580603438748e-06,
+      "loss": 0.8274,
+      "step": 7073
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6075153854514934,
+      "learning_rate": 3.8489827637213e-06,
+      "loss": 0.7427,
+      "step": 7074
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.764959273544147,
+      "learning_rate": 3.846385591600582e-06,
+      "loss": 0.7193,
+      "step": 7075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5309895025303242,
+      "learning_rate": 3.843789087358481e-06,
+      "loss": 0.6079,
+      "step": 7076
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7067092529866092,
+      "learning_rate": 3.841193251276797e-06,
+      "loss": 0.6646,
+      "step": 7077
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.54944657886839,
+      "learning_rate": 3.838598083637273e-06,
+      "loss": 0.598,
+      "step": 7078
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7348919530284566,
+      "learning_rate": 3.836003584721577e-06,
+      "loss": 0.7491,
+      "step": 7079
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.404024633603411,
+      "learning_rate": 3.833409754811295e-06,
+      "loss": 0.6814,
+      "step": 7080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.555957462060979,
+      "learning_rate": 3.830816594187951e-06,
+      "loss": 0.748,
+      "step": 7081
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7083144056582373,
+      "learning_rate": 3.828224103132989e-06,
+      "loss": 0.6971,
+      "step": 7082
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4838984338133978,
+      "learning_rate": 3.825632281927784e-06,
+      "loss": 0.6188,
+      "step": 7083
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6210945937368182,
+      "learning_rate": 3.823041130853642e-06,
+      "loss": 0.7154,
+      "step": 7084
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6502752504544569,
+      "learning_rate": 3.820450650191785e-06,
+      "loss": 0.609,
+      "step": 7085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8119556281486937,
+      "learning_rate": 3.817860840223373e-06,
+      "loss": 0.6798,
+      "step": 7086
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6083606169328408,
+      "learning_rate": 3.815271701229491e-06,
+      "loss": 0.6415,
+      "step": 7087
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7234006915737945,
+      "learning_rate": 3.8126832334911457e-06,
+      "loss": 0.7479,
+      "step": 7088
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7823470873162353,
+      "learning_rate": 3.8100954372892796e-06,
+      "loss": 0.7111,
+      "step": 7089
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.54738129409935,
+      "learning_rate": 3.8075083129047507e-06,
+      "loss": 0.6821,
+      "step": 7090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6814801668504775,
+      "learning_rate": 3.8049218606183547e-06,
+      "loss": 0.616,
+      "step": 7091
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8021188952408784,
+      "learning_rate": 3.8023360807108135e-06,
+      "loss": 0.7924,
+      "step": 7092
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.743377592740335,
+      "learning_rate": 3.799750973462768e-06,
+      "loss": 0.7971,
+      "step": 7093
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6937420380740584,
+      "learning_rate": 3.797166539154795e-06,
+      "loss": 0.6509,
+      "step": 7094
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6321361424032252,
+      "learning_rate": 3.7945827780673904e-06,
+      "loss": 0.6793,
+      "step": 7095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.603854582495709,
+      "learning_rate": 3.791999690480984e-06,
+      "loss": 0.6254,
+      "step": 7096
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.593723128503487,
+      "learning_rate": 3.7894172766759307e-06,
+      "loss": 0.6594,
+      "step": 7097
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5564792340475972,
+      "learning_rate": 3.786835536932507e-06,
+      "loss": 0.6598,
+      "step": 7098
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5831785482522764,
+      "learning_rate": 3.784254471530926e-06,
+      "loss": 0.7586,
+      "step": 7099
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7443724076857514,
+      "learning_rate": 3.7816740807513154e-06,
+      "loss": 0.6868,
+      "step": 7100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.509927809623996,
+      "learning_rate": 3.7790943648737397e-06,
+      "loss": 0.6671,
+      "step": 7101
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.725056179937763,
+      "learning_rate": 3.7765153241781893e-06,
+      "loss": 0.7384,
+      "step": 7102
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6151974285444082,
+      "learning_rate": 3.773936958944574e-06,
+      "loss": 0.649,
+      "step": 7103
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6428686899132872,
+      "learning_rate": 3.7713592694527357e-06,
+      "loss": 0.6798,
+      "step": 7104
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5815525541499031,
+      "learning_rate": 3.768782255982447e-06,
+      "loss": 0.6567,
+      "step": 7105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.693255150657801,
+      "learning_rate": 3.766205918813397e-06,
+      "loss": 0.6504,
+      "step": 7106
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7690595597513055,
+      "learning_rate": 3.7636302582252114e-06,
+      "loss": 0.7456,
+      "step": 7107
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.584515931711388,
+      "learning_rate": 3.761055274497433e-06,
+      "loss": 0.6839,
+      "step": 7108
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5816360333856814,
+      "learning_rate": 3.758480967909537e-06,
+      "loss": 0.6714,
+      "step": 7109
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.566909665599923,
+      "learning_rate": 3.7559073387409286e-06,
+      "loss": 0.7095,
+      "step": 7110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4996558972408467,
+      "learning_rate": 3.7533343872709295e-06,
+      "loss": 0.7148,
+      "step": 7111
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7782962742990462,
+      "learning_rate": 3.7507621137787988e-06,
+      "loss": 0.7736,
+      "step": 7112
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5960500136687246,
+      "learning_rate": 3.7481905185437105e-06,
+      "loss": 0.7787,
+      "step": 7113
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.606927467079132,
+      "learning_rate": 3.745619601844773e-06,
+      "loss": 0.6923,
+      "step": 7114
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.455797703919836,
+      "learning_rate": 3.7430493639610245e-06,
+      "loss": 0.6307,
+      "step": 7115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5626622983287617,
+      "learning_rate": 3.7404798051714165e-06,
+      "loss": 0.695,
+      "step": 7116
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6079084286137084,
+      "learning_rate": 3.7379109257548406e-06,
+      "loss": 0.7181,
+      "step": 7117
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4823136615497665,
+      "learning_rate": 3.7353427259901022e-06,
+      "loss": 0.7078,
+      "step": 7118
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6497796267623495,
+      "learning_rate": 3.732775206155943e-06,
+      "loss": 0.7823,
+      "step": 7119
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5963666093680742,
+      "learning_rate": 3.7302083665310295e-06,
+      "loss": 0.5863,
+      "step": 7120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7135779255987058,
+      "learning_rate": 3.7276422073939466e-06,
+      "loss": 0.7041,
+      "step": 7121
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.729275030874236,
+      "learning_rate": 3.7250767290232124e-06,
+      "loss": 0.7154,
+      "step": 7122
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5755392311483456,
+      "learning_rate": 3.7225119316972747e-06,
+      "loss": 0.6911,
+      "step": 7123
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.663429556378475,
+      "learning_rate": 3.7199478156944947e-06,
+      "loss": 0.6331,
+      "step": 7124
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.711823333115215,
+      "learning_rate": 3.717384381293173e-06,
+      "loss": 0.7237,
+      "step": 7125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6672270773065978,
+      "learning_rate": 3.7148216287715243e-06,
+      "loss": 0.7352,
+      "step": 7126
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8010218071305932,
+      "learning_rate": 3.712259558407698e-06,
+      "loss": 0.638,
+      "step": 7127
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7201988979672762,
+      "learning_rate": 3.709698170479772e-06,
+      "loss": 0.6107,
+      "step": 7128
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7275487216334133,
+      "learning_rate": 3.7071374652657366e-06,
+      "loss": 0.8083,
+      "step": 7129
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6241632422864194,
+      "learning_rate": 3.704577443043523e-06,
+      "loss": 0.6765,
+      "step": 7130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5499977286232118,
+      "learning_rate": 3.7020181040909755e-06,
+      "loss": 0.6343,
+      "step": 7131
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6018603460961796,
+      "learning_rate": 3.6994594486858724e-06,
+      "loss": 0.6779,
+      "step": 7132
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6908374645330553,
+      "learning_rate": 3.696901477105921e-06,
+      "loss": 0.6236,
+      "step": 7133
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.521919889461445,
+      "learning_rate": 3.694344189628741e-06,
+      "loss": 0.6738,
+      "step": 7134
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7491997799433427,
+      "learning_rate": 3.6917875865318944e-06,
+      "loss": 0.6773,
+      "step": 7135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6868041845930088,
+      "learning_rate": 3.6892316680928508e-06,
+      "loss": 0.6554,
+      "step": 7136
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.556084069113492,
+      "learning_rate": 3.6866764345890217e-06,
+      "loss": 0.6641,
+      "step": 7137
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4191396117041615,
+      "learning_rate": 3.684121886297739e-06,
+      "loss": 0.6768,
+      "step": 7138
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6062587399945396,
+      "learning_rate": 3.6815680234962527e-06,
+      "loss": 0.5832,
+      "step": 7139
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7773274881370058,
+      "learning_rate": 3.6790148464617483e-06,
+      "loss": 0.7483,
+      "step": 7140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7298207064959892,
+      "learning_rate": 3.6764623554713364e-06,
+      "loss": 0.7185,
+      "step": 7141
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.802729381434695,
+      "learning_rate": 3.6739105508020433e-06,
+      "loss": 0.7059,
+      "step": 7142
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.55814796727082,
+      "learning_rate": 3.6713594327308343e-06,
+      "loss": 0.6882,
+      "step": 7143
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9841458785651938,
+      "learning_rate": 3.6688090015345878e-06,
+      "loss": 0.7185,
+      "step": 7144
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5236445095995443,
+      "learning_rate": 3.6662592574901147e-06,
+      "loss": 0.763,
+      "step": 7145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.667091898622156,
+      "learning_rate": 3.663710200874153e-06,
+      "loss": 0.6871,
+      "step": 7146
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.640283370489685,
+      "learning_rate": 3.661161831963358e-06,
+      "loss": 0.6242,
+      "step": 7147
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5959959058927706,
+      "learning_rate": 3.658614151034321e-06,
+      "loss": 0.7617,
+      "step": 7148
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7141601341216794,
+      "learning_rate": 3.6560671583635467e-06,
+      "loss": 0.6534,
+      "step": 7149
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7300259568139287,
+      "learning_rate": 3.653520854227475e-06,
+      "loss": 0.7433,
+      "step": 7150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5532886686274163,
+      "learning_rate": 3.6509752389024687e-06,
+      "loss": 0.579,
+      "step": 7151
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6833266687520791,
+      "learning_rate": 3.648430312664809e-06,
+      "loss": 0.6755,
+      "step": 7152
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5564738321146763,
+      "learning_rate": 3.6458860757907155e-06,
+      "loss": 0.6367,
+      "step": 7153
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7105058056686722,
+      "learning_rate": 3.6433425285563186e-06,
+      "loss": 0.6857,
+      "step": 7154
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6487205971867094,
+      "learning_rate": 3.6407996712376826e-06,
+      "loss": 0.6234,
+      "step": 7155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8336973044214955,
+      "learning_rate": 3.6382575041107982e-06,
+      "loss": 0.735,
+      "step": 7156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6426974325555639,
+      "learning_rate": 3.6357160274515723e-06,
+      "loss": 0.6363,
+      "step": 7157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5300249782218753,
+      "learning_rate": 3.633175241535849e-06,
+      "loss": 0.6511,
+      "step": 7158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6712026210826763,
+      "learning_rate": 3.630635146639384e-06,
+      "loss": 0.7023,
+      "step": 7159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5512216505261869,
+      "learning_rate": 3.6280957430378683e-06,
+      "loss": 0.6601,
+      "step": 7160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6199132742303428,
+      "learning_rate": 3.6255570310069166e-06,
+      "loss": 0.6928,
+      "step": 7161
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7001832013414149,
+      "learning_rate": 3.6230190108220607e-06,
+      "loss": 0.7222,
+      "step": 7162
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5664502892848486,
+      "learning_rate": 3.6204816827587676e-06,
+      "loss": 0.6602,
+      "step": 7163
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7106532424063383,
+      "learning_rate": 3.617945047092426e-06,
+      "loss": 0.6976,
+      "step": 7164
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5093380311405362,
+      "learning_rate": 3.615409104098342e-06,
+      "loss": 0.5972,
+      "step": 7165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6008320342553601,
+      "learning_rate": 3.612873854051759e-06,
+      "loss": 0.6195,
+      "step": 7166
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.65846303644904,
+      "learning_rate": 3.610339297227833e-06,
+      "loss": 0.7601,
+      "step": 7167
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6312980744454015,
+      "learning_rate": 3.6078054339016523e-06,
+      "loss": 0.7859,
+      "step": 7168
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6668920120325388,
+      "learning_rate": 3.605272264348233e-06,
+      "loss": 0.697,
+      "step": 7169
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.581682435271611,
+      "learning_rate": 3.602739788842503e-06,
+      "loss": 0.6864,
+      "step": 7170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.622319941173912,
+      "learning_rate": 3.6002080076593273e-06,
+      "loss": 0.732,
+      "step": 7171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6851021641033066,
+      "learning_rate": 3.597676921073494e-06,
+      "loss": 0.7444,
+      "step": 7172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7772998980522972,
+      "learning_rate": 3.5951465293597067e-06,
+      "loss": 0.7153,
+      "step": 7173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7090383637910858,
+      "learning_rate": 3.5926168327926037e-06,
+      "loss": 0.7717,
+      "step": 7174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5283316067335495,
+      "learning_rate": 3.590087831646746e-06,
+      "loss": 0.5902,
+      "step": 7175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.70593382715811,
+      "learning_rate": 3.58755952619661e-06,
+      "loss": 0.7558,
+      "step": 7176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4769021537695464,
+      "learning_rate": 3.585031916716609e-06,
+      "loss": 0.5626,
+      "step": 7177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5193009552710477,
+      "learning_rate": 3.5825050034810783e-06,
+      "loss": 0.6463,
+      "step": 7178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5464268469172968,
+      "learning_rate": 3.5799787867642668e-06,
+      "loss": 0.6902,
+      "step": 7179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9070860927402575,
+      "learning_rate": 3.5774532668403606e-06,
+      "loss": 0.8036,
+      "step": 7180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.677251422282328,
+      "learning_rate": 3.5749284439834686e-06,
+      "loss": 0.7145,
+      "step": 7181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5070874738175213,
+      "learning_rate": 3.5724043184676127e-06,
+      "loss": 0.6111,
+      "step": 7182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3679124163375487,
+      "learning_rate": 3.5698808905667525e-06,
+      "loss": 0.6639,
+      "step": 7183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6294192752826915,
+      "learning_rate": 3.567358160554769e-06,
+      "loss": 0.6482,
+      "step": 7184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6115377083728697,
+      "learning_rate": 3.564836128705459e-06,
+      "loss": 0.6533,
+      "step": 7185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9367670221869318,
+      "learning_rate": 3.562314795292552e-06,
+      "loss": 0.6958,
+      "step": 7186
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5439196900733712,
+      "learning_rate": 3.5597941605897037e-06,
+      "loss": 0.6759,
+      "step": 7187
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8954483496019165,
+      "learning_rate": 3.5572742248704817e-06,
+      "loss": 0.702,
+      "step": 7188
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5821038172561497,
+      "learning_rate": 3.554754988408393e-06,
+      "loss": 0.5861,
+      "step": 7189
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6938894940828406,
+      "learning_rate": 3.552236451476855e-06,
+      "loss": 0.7871,
+      "step": 7190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.614556870374976,
+      "learning_rate": 3.5497186143492178e-06,
+      "loss": 0.7146,
+      "step": 7191
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5828409470899927,
+      "learning_rate": 3.5472014772987573e-06,
+      "loss": 0.7527,
+      "step": 7192
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6435787132195512,
+      "learning_rate": 3.544685040598663e-06,
+      "loss": 0.6479,
+      "step": 7193
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7080262144710854,
+      "learning_rate": 3.5421693045220617e-06,
+      "loss": 0.6583,
+      "step": 7194
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6662616193281345,
+      "learning_rate": 3.539654269341989e-06,
+      "loss": 0.7191,
+      "step": 7195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7599616953319472,
+      "learning_rate": 3.5371399353314174e-06,
+      "loss": 0.659,
+      "step": 7196
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5910727310443404,
+      "learning_rate": 3.534626302763241e-06,
+      "loss": 0.6978,
+      "step": 7197
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7541929338966649,
+      "learning_rate": 3.5321133719102685e-06,
+      "loss": 0.665,
+      "step": 7198
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.563447514004173,
+      "learning_rate": 3.529601143045244e-06,
+      "loss": 0.5738,
+      "step": 7199
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6272858485757566,
+      "learning_rate": 3.527089616440832e-06,
+      "loss": 0.6693,
+      "step": 7200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6767842031362263,
+      "learning_rate": 3.524578792369614e-06,
+      "loss": 0.6784,
+      "step": 7201
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.741601124837262,
+      "learning_rate": 3.5220686711041085e-06,
+      "loss": 0.6576,
+      "step": 7202
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6554235824440988,
+      "learning_rate": 3.51955925291674e-06,
+      "loss": 0.7596,
+      "step": 7203
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8480889756347079,
+      "learning_rate": 3.5170505380798736e-06,
+      "loss": 0.7017,
+      "step": 7204
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7918181666049604,
+      "learning_rate": 3.5145425268657916e-06,
+      "loss": 0.7029,
+      "step": 7205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.790400059060796,
+      "learning_rate": 3.5120352195466932e-06,
+      "loss": 0.6969,
+      "step": 7206
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6471726588298956,
+      "learning_rate": 3.509528616394716e-06,
+      "loss": 0.6233,
+      "step": 7207
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4980354612253914,
+      "learning_rate": 3.5070227176819036e-06,
+      "loss": 0.7079,
+      "step": 7208
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.570210483481177,
+      "learning_rate": 3.5045175236802363e-06,
+      "loss": 0.7148,
+      "step": 7209
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.553566799177608,
+      "learning_rate": 3.502013034661619e-06,
+      "loss": 0.6731,
+      "step": 7210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8000474948425484,
+      "learning_rate": 3.4995092508978647e-06,
+      "loss": 0.7375,
+      "step": 7211
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6913024060790467,
+      "learning_rate": 3.4970061726607273e-06,
+      "loss": 0.6878,
+      "step": 7212
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7068333067586372,
+      "learning_rate": 3.494503800221872e-06,
+      "loss": 0.71,
+      "step": 7213
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.817682464669328,
+      "learning_rate": 3.4920021338528955e-06,
+      "loss": 0.672,
+      "step": 7214
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6311735296816936,
+      "learning_rate": 3.489501173825316e-06,
+      "loss": 0.7779,
+      "step": 7215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.50529393797726,
+      "learning_rate": 3.487000920410568e-06,
+      "loss": 0.6823,
+      "step": 7216
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6558426245962694,
+      "learning_rate": 3.4845013738800194e-06,
+      "loss": 0.61,
+      "step": 7217
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.612431482351146,
+      "learning_rate": 3.482002534504958e-06,
+      "loss": 0.7768,
+      "step": 7218
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7138504683026008,
+      "learning_rate": 3.479504402556587e-06,
+      "loss": 0.6834,
+      "step": 7219
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5804252165412198,
+      "learning_rate": 3.477006978306049e-06,
+      "loss": 0.7372,
+      "step": 7220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5959790428928604,
+      "learning_rate": 3.474510262024391e-06,
+      "loss": 0.6769,
+      "step": 7221
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6203807255279645,
+      "learning_rate": 3.4720142539825975e-06,
+      "loss": 0.691,
+      "step": 7222
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7315453441096231,
+      "learning_rate": 3.4695189544515727e-06,
+      "loss": 0.6951,
+      "step": 7223
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.508462206764295,
+      "learning_rate": 3.467024363702138e-06,
+      "loss": 0.6509,
+      "step": 7224
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7963743245623451,
+      "learning_rate": 3.4645304820050473e-06,
+      "loss": 0.6806,
+      "step": 7225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.538399822037882,
+      "learning_rate": 3.4620373096309656e-06,
+      "loss": 0.6781,
+      "step": 7226
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7490947649773623,
+      "learning_rate": 3.459544846850492e-06,
+      "loss": 0.6844,
+      "step": 7227
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7330712725230233,
+      "learning_rate": 3.4570530939341472e-06,
+      "loss": 0.6949,
+      "step": 7228
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6639851511795398,
+      "learning_rate": 3.454562051152366e-06,
+      "loss": 0.7073,
+      "step": 7229
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.667892739832757,
+      "learning_rate": 3.452071718775518e-06,
+      "loss": 0.7281,
+      "step": 7230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5559555783691887,
+      "learning_rate": 3.4495820970738834e-06,
+      "loss": 0.6977,
+      "step": 7231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8376998130377118,
+      "learning_rate": 3.4470931863176747e-06,
+      "loss": 0.6824,
+      "step": 7232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.624520742992479,
+      "learning_rate": 3.4446049867770283e-06,
+      "loss": 0.642,
+      "step": 7233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.665652967735254,
+      "learning_rate": 3.442117498721992e-06,
+      "loss": 0.6934,
+      "step": 7234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.495900862109376,
+      "learning_rate": 3.439630722422548e-06,
+      "loss": 0.6932,
+      "step": 7235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6200504827919735,
+      "learning_rate": 3.437144658148599e-06,
+      "loss": 0.6874,
+      "step": 7236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6871382210453179,
+      "learning_rate": 3.434659306169963e-06,
+      "loss": 0.7193,
+      "step": 7237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6020384570556934,
+      "learning_rate": 3.4321746667563925e-06,
+      "loss": 0.6491,
+      "step": 7238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4852594541553965,
+      "learning_rate": 3.4296907401775494e-06,
+      "loss": 0.612,
+      "step": 7239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.653249261782761,
+      "learning_rate": 3.4272075267030267e-06,
+      "loss": 0.6335,
+      "step": 7240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5842397819032858,
+      "learning_rate": 3.4247250266023446e-06,
+      "loss": 0.6908,
+      "step": 7241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5817026113440948,
+      "learning_rate": 3.4222432401449313e-06,
+      "loss": 0.7081,
+      "step": 7242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6773022328073637,
+      "learning_rate": 3.4197621676001537e-06,
+      "loss": 0.7217,
+      "step": 7243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.582428050031458,
+      "learning_rate": 3.417281809237285e-06,
+      "loss": 0.6992,
+      "step": 7244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6711688415393713,
+      "learning_rate": 3.4148021653255347e-06,
+      "loss": 0.754,
+      "step": 7245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8417283173332568,
+      "learning_rate": 3.412323236134032e-06,
+      "loss": 0.6941,
+      "step": 7246
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8817162021377392,
+      "learning_rate": 3.409845021931818e-06,
+      "loss": 0.695,
+      "step": 7247
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5693407212436574,
+      "learning_rate": 3.4073675229878724e-06,
+      "loss": 0.6799,
+      "step": 7248
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.602686936513114,
+      "learning_rate": 3.404890739571081e-06,
+      "loss": 0.7407,
+      "step": 7249
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5627001624359027,
+      "learning_rate": 3.4024146719502636e-06,
+      "loss": 0.6668,
+      "step": 7250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9331849062270565,
+      "learning_rate": 3.3999393203941623e-06,
+      "loss": 0.731,
+      "step": 7251
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5206783627275,
+      "learning_rate": 3.397464685171431e-06,
+      "loss": 0.6608,
+      "step": 7252
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7107663982664039,
+      "learning_rate": 3.3949907665506564e-06,
+      "loss": 0.73,
+      "step": 7253
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8654410443640381,
+      "learning_rate": 3.3925175648003473e-06,
+      "loss": 0.7268,
+      "step": 7254
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6870964608863164,
+      "learning_rate": 3.390045080188923e-06,
+      "loss": 0.6921,
+      "step": 7255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6475194906845443,
+      "learning_rate": 3.38757331298474e-06,
+      "loss": 0.7748,
+      "step": 7256
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8213267285460604,
+      "learning_rate": 3.385102263456065e-06,
+      "loss": 0.5969,
+      "step": 7257
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6459408638945416,
+      "learning_rate": 3.3826319318710933e-06,
+      "loss": 0.697,
+      "step": 7258
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.703788221164889,
+      "learning_rate": 3.3801623184979437e-06,
+      "loss": 0.6976,
+      "step": 7259
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7504852878801893,
+      "learning_rate": 3.377693423604649e-06,
+      "loss": 0.7039,
+      "step": 7260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5253648258267434,
+      "learning_rate": 3.3752252474591762e-06,
+      "loss": 0.6294,
+      "step": 7261
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5040219167465418,
+      "learning_rate": 3.3727577903293985e-06,
+      "loss": 0.631,
+      "step": 7262
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7870140766431388,
+      "learning_rate": 3.3702910524831246e-06,
+      "loss": 0.7444,
+      "step": 7263
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.607093992786678,
+      "learning_rate": 3.3678250341880825e-06,
+      "loss": 0.7117,
+      "step": 7264
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5595706608009965,
+      "learning_rate": 3.365359735711915e-06,
+      "loss": 0.6255,
+      "step": 7265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7040629735909656,
+      "learning_rate": 3.3628951573221966e-06,
+      "loss": 0.7388,
+      "step": 7266
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7937033138858276,
+      "learning_rate": 3.360431299286413e-06,
+      "loss": 0.6807,
+      "step": 7267
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6216251520551597,
+      "learning_rate": 3.3579681618719805e-06,
+      "loss": 0.7217,
+      "step": 7268
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5615083234069798,
+      "learning_rate": 3.355505745346238e-06,
+      "loss": 0.6231,
+      "step": 7269
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.532634334500432,
+      "learning_rate": 3.353044049976435e-06,
+      "loss": 0.645,
+      "step": 7270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.638134421357406,
+      "learning_rate": 3.3505830760297543e-06,
+      "loss": 0.6466,
+      "step": 7271
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.653769742345302,
+      "learning_rate": 3.3481228237732975e-06,
+      "loss": 0.7258,
+      "step": 7272
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5837947311775986,
+      "learning_rate": 3.345663293474083e-06,
+      "loss": 0.6474,
+      "step": 7273
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.64849332588951,
+      "learning_rate": 3.3432044853990576e-06,
+      "loss": 0.6081,
+      "step": 7274
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6257045316549825,
+      "learning_rate": 3.3407463998150815e-06,
+      "loss": 0.5885,
+      "step": 7275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9181770773628148,
+      "learning_rate": 3.338289036988945e-06,
+      "loss": 0.6925,
+      "step": 7276
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6579765006359775,
+      "learning_rate": 3.3358323971873586e-06,
+      "loss": 0.7907,
+      "step": 7277
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6223788293680472,
+      "learning_rate": 3.3333764806769474e-06,
+      "loss": 0.7057,
+      "step": 7278
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5562895761921027,
+      "learning_rate": 3.3309212877242635e-06,
+      "loss": 0.6549,
+      "step": 7279
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8141933918317033,
+      "learning_rate": 3.3284668185957857e-06,
+      "loss": 0.7398,
+      "step": 7280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.54263284558655,
+      "learning_rate": 3.3260130735579e-06,
+      "loss": 0.6571,
+      "step": 7281
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.619131530660715,
+      "learning_rate": 3.3235600528769253e-06,
+      "loss": 0.7058,
+      "step": 7282
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.465021228430089,
+      "learning_rate": 3.321107756819103e-06,
+      "loss": 0.5369,
+      "step": 7283
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6381336452278406,
+      "learning_rate": 3.318656185650584e-06,
+      "loss": 0.6229,
+      "step": 7284
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5536313913242494,
+      "learning_rate": 3.3162053396374504e-06,
+      "loss": 0.6387,
+      "step": 7285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6618093267828662,
+      "learning_rate": 3.313755219045709e-06,
+      "loss": 0.6368,
+      "step": 7286
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6517390858077126,
+      "learning_rate": 3.311305824141273e-06,
+      "loss": 0.6692,
+      "step": 7287
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4256680556420662,
+      "learning_rate": 3.308857155189992e-06,
+      "loss": 0.7384,
+      "step": 7288
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5064498910789945,
+      "learning_rate": 3.306409212457631e-06,
+      "loss": 0.6581,
+      "step": 7289
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7608860865876834,
+      "learning_rate": 3.30396199620987e-06,
+      "loss": 0.6862,
+      "step": 7290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6037617675529898,
+      "learning_rate": 3.301515506712322e-06,
+      "loss": 0.7146,
+      "step": 7291
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7593652994874995,
+      "learning_rate": 3.2990697442305143e-06,
+      "loss": 0.6931,
+      "step": 7292
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5088961297220562,
+      "learning_rate": 3.296624709029893e-06,
+      "loss": 0.5407,
+      "step": 7293
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.549978531019286,
+      "learning_rate": 3.2941804013758337e-06,
+      "loss": 0.7049,
+      "step": 7294
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6325363025182515,
+      "learning_rate": 3.291736821533621e-06,
+      "loss": 0.7165,
+      "step": 7295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6735142078675114,
+      "learning_rate": 3.289293969768471e-06,
+      "loss": 0.6012,
+      "step": 7296
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7178665383223046,
+      "learning_rate": 3.2868518463455203e-06,
+      "loss": 0.6026,
+      "step": 7297
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6928985980299383,
+      "learning_rate": 3.284410451529816e-06,
+      "loss": 0.65,
+      "step": 7298
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7305481371698612,
+      "learning_rate": 3.2819697855863374e-06,
+      "loss": 0.7129,
+      "step": 7299
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7258542451488783,
+      "learning_rate": 3.279529848779983e-06,
+      "loss": 0.6139,
+      "step": 7300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6523221761729956,
+      "learning_rate": 3.2770906413755644e-06,
+      "loss": 0.7421,
+      "step": 7301
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.534840481965886,
+      "learning_rate": 3.274652163637826e-06,
+      "loss": 0.6207,
+      "step": 7302
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7502912095179775,
+      "learning_rate": 3.2722144158314184e-06,
+      "loss": 0.7288,
+      "step": 7303
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6496887127591713,
+      "learning_rate": 3.2697773982209257e-06,
+      "loss": 0.6785,
+      "step": 7304
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6368353847674915,
+      "learning_rate": 3.2673411110708507e-06,
+      "loss": 0.666,
+      "step": 7305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5451756159890135,
+      "learning_rate": 3.2649055546456076e-06,
+      "loss": 0.7173,
+      "step": 7306
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7004419330703944,
+      "learning_rate": 3.2624707292095446e-06,
+      "loss": 0.7743,
+      "step": 7307
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.595740560798837,
+      "learning_rate": 3.2600366350269176e-06,
+      "loss": 0.5945,
+      "step": 7308
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.722661566376523,
+      "learning_rate": 3.257603272361912e-06,
+      "loss": 0.6278,
+      "step": 7309
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6681602626098242,
+      "learning_rate": 3.2551706414786366e-06,
+      "loss": 0.6901,
+      "step": 7310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6689875835667756,
+      "learning_rate": 3.252738742641106e-06,
+      "loss": 0.7482,
+      "step": 7311
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6118580407253362,
+      "learning_rate": 3.2503075761132707e-06,
+      "loss": 0.6896,
+      "step": 7312
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6008888195431321,
+      "learning_rate": 3.247877142158997e-06,
+      "loss": 0.7138,
+      "step": 7313
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6096208333217168,
+      "learning_rate": 3.2454474410420657e-06,
+      "loss": 0.6315,
+      "step": 7314
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6230174119350946,
+      "learning_rate": 3.2430184730261883e-06,
+      "loss": 0.6744,
+      "step": 7315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.685018707256806,
+      "learning_rate": 3.2405902383749856e-06,
+      "loss": 0.684,
+      "step": 7316
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6803874750251426,
+      "learning_rate": 3.2381627373520066e-06,
+      "loss": 0.7465,
+      "step": 7317
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6429226662638055,
+      "learning_rate": 3.2357359702207216e-06,
+      "loss": 0.7916,
+      "step": 7318
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7563407976090484,
+      "learning_rate": 3.233309937244513e-06,
+      "loss": 0.736,
+      "step": 7319
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5215856115164546,
+      "learning_rate": 3.2308846386866954e-06,
+      "loss": 0.6469,
+      "step": 7320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5886452708615701,
+      "learning_rate": 3.228460074810489e-06,
+      "loss": 0.7408,
+      "step": 7321
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6400376738359026,
+      "learning_rate": 3.226036245879047e-06,
+      "loss": 0.6373,
+      "step": 7322
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6690055408195899,
+      "learning_rate": 3.2236131521554405e-06,
+      "loss": 0.6703,
+      "step": 7323
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5373330223676502,
+      "learning_rate": 3.221190793902653e-06,
+      "loss": 0.5949,
+      "step": 7324
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6379524975379682,
+      "learning_rate": 3.218769171383599e-06,
+      "loss": 0.6989,
+      "step": 7325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.819067342413871,
+      "learning_rate": 3.2163482848611003e-06,
+      "loss": 0.6376,
+      "step": 7326
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7154236629745496,
+      "learning_rate": 3.213928134597912e-06,
+      "loss": 0.7656,
+      "step": 7327
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6693865924740787,
+      "learning_rate": 3.2115087208567064e-06,
+      "loss": 0.7179,
+      "step": 7328
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7508571742865737,
+      "learning_rate": 3.209090043900065e-06,
+      "loss": 0.696,
+      "step": 7329
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5057491696391703,
+      "learning_rate": 3.206672103990501e-06,
+      "loss": 0.596,
+      "step": 7330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6512693921958572,
+      "learning_rate": 3.2042549013904476e-06,
+      "loss": 0.7441,
+      "step": 7331
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7628128435052854,
+      "learning_rate": 3.201838436362249e-06,
+      "loss": 0.6603,
+      "step": 7332
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6365190818788702,
+      "learning_rate": 3.1994227091681783e-06,
+      "loss": 0.6335,
+      "step": 7333
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5780363922604819,
+      "learning_rate": 3.197007720070422e-06,
+      "loss": 0.6621,
+      "step": 7334
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.572747001069289,
+      "learning_rate": 3.1945934693310897e-06,
+      "loss": 0.6726,
+      "step": 7335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5666495089845731,
+      "learning_rate": 3.192179957212215e-06,
+      "loss": 0.6427,
+      "step": 7336
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6216586270269608,
+      "learning_rate": 3.1897671839757405e-06,
+      "loss": 0.6585,
+      "step": 7337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5232294696713657,
+      "learning_rate": 3.187355149883541e-06,
+      "loss": 0.6636,
+      "step": 7338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5701600681094992,
+      "learning_rate": 3.1849438551973986e-06,
+      "loss": 0.6563,
+      "step": 7339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7231236105797698,
+      "learning_rate": 3.182533300179026e-06,
+      "loss": 0.672,
+      "step": 7340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.660836767656615,
+      "learning_rate": 3.180123485090053e-06,
+      "loss": 0.6324,
+      "step": 7341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6430898224518655,
+      "learning_rate": 3.177714410192022e-06,
+      "loss": 0.6109,
+      "step": 7342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6269418239884736,
+      "learning_rate": 3.175306075746406e-06,
+      "loss": 0.7617,
+      "step": 7343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6439427634892574,
+      "learning_rate": 3.172898482014586e-06,
+      "loss": 0.6748,
+      "step": 7344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7691958231320482,
+      "learning_rate": 3.1704916292578715e-06,
+      "loss": 0.6396,
+      "step": 7345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.637905279936545,
+      "learning_rate": 3.1680855177374935e-06,
+      "loss": 0.6897,
+      "step": 7346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6798196456772034,
+      "learning_rate": 3.1656801477145893e-06,
+      "loss": 0.7228,
+      "step": 7347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7038195826667484,
+      "learning_rate": 3.1632755194502286e-06,
+      "loss": 0.6807,
+      "step": 7348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6980391062663183,
+      "learning_rate": 3.160871633205398e-06,
+      "loss": 0.6299,
+      "step": 7349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4863795099644443,
+      "learning_rate": 3.1584684892409977e-06,
+      "loss": 0.5528,
+      "step": 7350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5142489059614048,
+      "learning_rate": 3.156066087817856e-06,
+      "loss": 0.6406,
+      "step": 7351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7486622198807542,
+      "learning_rate": 3.15366442919671e-06,
+      "loss": 0.6327,
+      "step": 7352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6055617758275391,
+      "learning_rate": 3.1512635136382252e-06,
+      "loss": 0.5865,
+      "step": 7353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5219472276420638,
+      "learning_rate": 3.1488633414029877e-06,
+      "loss": 0.6461,
+      "step": 7354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8176492653029488,
+      "learning_rate": 3.146463912751491e-06,
+      "loss": 0.7689,
+      "step": 7355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6826159035851425,
+      "learning_rate": 3.144065227944163e-06,
+      "loss": 0.7403,
+      "step": 7356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6247469531612129,
+      "learning_rate": 3.1416672872413357e-06,
+      "loss": 0.7106,
+      "step": 7357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5779106715212203,
+      "learning_rate": 3.1392700909032724e-06,
+      "loss": 0.6871,
+      "step": 7358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4920076107625604,
+      "learning_rate": 3.1368736391901544e-06,
+      "loss": 0.6714,
+      "step": 7359
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.961808390580555,
+      "learning_rate": 3.134477932362072e-06,
+      "loss": 0.7265,
+      "step": 7360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.938920841767802,
+      "learning_rate": 3.132082970679049e-06,
+      "loss": 0.7427,
+      "step": 7361
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.567752759558388,
+      "learning_rate": 3.1296887544010133e-06,
+      "loss": 0.6539,
+      "step": 7362
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5151465329107645,
+      "learning_rate": 3.1272952837878245e-06,
+      "loss": 0.6595,
+      "step": 7363
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.705679499416008,
+      "learning_rate": 3.124902559099258e-06,
+      "loss": 0.6997,
+      "step": 7364
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8858540122131102,
+      "learning_rate": 3.1225105805950017e-06,
+      "loss": 0.7913,
+      "step": 7365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7419499972808716,
+      "learning_rate": 3.120119348534669e-06,
+      "loss": 0.7856,
+      "step": 7366
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.573465112824827,
+      "learning_rate": 3.1177288631777956e-06,
+      "loss": 0.6733,
+      "step": 7367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6325035220696849,
+      "learning_rate": 3.115339124783825e-06,
+      "loss": 0.6511,
+      "step": 7368
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7307285220036097,
+      "learning_rate": 3.1129501336121303e-06,
+      "loss": 0.615,
+      "step": 7369
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9298849037539572,
+      "learning_rate": 3.1105618899219946e-06,
+      "loss": 0.6952,
+      "step": 7370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8535262033904598,
+      "learning_rate": 3.1081743939726262e-06,
+      "loss": 0.7454,
+      "step": 7371
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6615890889245795,
+      "learning_rate": 3.105787646023155e-06,
+      "loss": 0.803,
+      "step": 7372
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7370173676111043,
+      "learning_rate": 3.103401646332619e-06,
+      "loss": 0.6846,
+      "step": 7373
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.810281054365534,
+      "learning_rate": 3.101016395159986e-06,
+      "loss": 0.7291,
+      "step": 7374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6837037222155278,
+      "learning_rate": 3.0986318927641312e-06,
+      "loss": 0.7795,
+      "step": 7375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7021722264339456,
+      "learning_rate": 3.09624813940386e-06,
+      "loss": 0.6884,
+      "step": 7376
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6763761247294078,
+      "learning_rate": 3.0938651353378946e-06,
+      "loss": 0.7276,
+      "step": 7377
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5836681850071694,
+      "learning_rate": 3.091482880824864e-06,
+      "loss": 0.6714,
+      "step": 7378
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.731965481202213,
+      "learning_rate": 3.0891013761233345e-06,
+      "loss": 0.6925,
+      "step": 7379
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7464462818615787,
+      "learning_rate": 3.0867206214917723e-06,
+      "loss": 0.7193,
+      "step": 7380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.849544297503174,
+      "learning_rate": 3.0843406171885746e-06,
+      "loss": 0.8064,
+      "step": 7381
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.540043474640889,
+      "learning_rate": 3.081961363472058e-06,
+      "loss": 0.7355,
+      "step": 7382
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6909442617771338,
+      "learning_rate": 3.0795828606004453e-06,
+      "loss": 0.6626,
+      "step": 7383
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7735932876842673,
+      "learning_rate": 3.0772051088318912e-06,
+      "loss": 0.6367,
+      "step": 7384
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6463646066423563,
+      "learning_rate": 3.0748281084244635e-06,
+      "loss": 0.7147,
+      "step": 7385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5406094829194776,
+      "learning_rate": 3.0724518596361452e-06,
+      "loss": 0.6248,
+      "step": 7386
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6502764101119622,
+      "learning_rate": 3.070076362724843e-06,
+      "loss": 0.5867,
+      "step": 7387
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4455671784961945,
+      "learning_rate": 3.0677016179483823e-06,
+      "loss": 0.7,
+      "step": 7388
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7474920723140612,
+      "learning_rate": 3.065327625564498e-06,
+      "loss": 0.6584,
+      "step": 7389
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6490934938705888,
+      "learning_rate": 3.0629543858308552e-06,
+      "loss": 0.6237,
+      "step": 7390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.620282671271724,
+      "learning_rate": 3.0605818990050327e-06,
+      "loss": 0.6551,
+      "step": 7391
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6635610737609734,
+      "learning_rate": 3.0582101653445227e-06,
+      "loss": 0.6701,
+      "step": 7392
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7532758162200444,
+      "learning_rate": 3.0558391851067416e-06,
+      "loss": 0.783,
+      "step": 7393
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6603614563648805,
+      "learning_rate": 3.053468958549025e-06,
+      "loss": 0.6462,
+      "step": 7394
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5306270774029103,
+      "learning_rate": 3.051099485928618e-06,
+      "loss": 0.6622,
+      "step": 7395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.500571714029283,
+      "learning_rate": 3.0487307675026967e-06,
+      "loss": 0.631,
+      "step": 7396
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3669527221428746,
+      "learning_rate": 3.0463628035283423e-06,
+      "loss": 0.6113,
+      "step": 7397
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6656957768204146,
+      "learning_rate": 3.043995594262562e-06,
+      "loss": 0.612,
+      "step": 7398
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7199610639093554,
+      "learning_rate": 3.0416291399622834e-06,
+      "loss": 0.7657,
+      "step": 7399
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5742384138792822,
+      "learning_rate": 3.039263440884341e-06,
+      "loss": 0.6016,
+      "step": 7400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5696383649368324,
+      "learning_rate": 3.0368984972854997e-06,
+      "loss": 0.5948,
+      "step": 7401
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5431594277390508,
+      "learning_rate": 3.034534309422439e-06,
+      "loss": 0.7673,
+      "step": 7402
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7611851032651105,
+      "learning_rate": 3.0321708775517476e-06,
+      "loss": 0.7376,
+      "step": 7403
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6782638099138119,
+      "learning_rate": 3.0298082019299424e-06,
+      "loss": 0.6546,
+      "step": 7404
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.655260128106923,
+      "learning_rate": 3.0274462828134587e-06,
+      "loss": 0.8029,
+      "step": 7405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7300337613368677,
+      "learning_rate": 3.0250851204586394e-06,
+      "loss": 0.64,
+      "step": 7406
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6253057086422333,
+      "learning_rate": 3.0227247151217553e-06,
+      "loss": 0.6651,
+      "step": 7407
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7339750324433152,
+      "learning_rate": 3.0203650670589945e-06,
+      "loss": 0.672,
+      "step": 7408
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.477359102111727,
+      "learning_rate": 3.0180061765264523e-06,
+      "loss": 0.658,
+      "step": 7409
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.621876733888337,
+      "learning_rate": 3.015648043780156e-06,
+      "loss": 0.5895,
+      "step": 7410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6668748997638514,
+      "learning_rate": 3.0132906690760398e-06,
+      "loss": 0.6339,
+      "step": 7411
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6796738882866564,
+      "learning_rate": 3.0109340526699604e-06,
+      "loss": 0.5918,
+      "step": 7412
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5084714791846123,
+      "learning_rate": 3.0085781948176974e-06,
+      "loss": 0.6234,
+      "step": 7413
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7944341554365648,
+      "learning_rate": 3.006223095774933e-06,
+      "loss": 0.7622,
+      "step": 7414
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7037480514842924,
+      "learning_rate": 3.003868755797286e-06,
+      "loss": 0.7089,
+      "step": 7415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7032944351810269,
+      "learning_rate": 3.0015151751402725e-06,
+      "loss": 0.6763,
+      "step": 7416
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.745469835993225,
+      "learning_rate": 2.9991623540593453e-06,
+      "loss": 0.71,
+      "step": 7417
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6436550601102142,
+      "learning_rate": 2.996810292809865e-06,
+      "loss": 0.8204,
+      "step": 7418
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6882237970692697,
+      "learning_rate": 2.994458991647108e-06,
+      "loss": 0.6516,
+      "step": 7419
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6759976105871366,
+      "learning_rate": 2.9921084508262722e-06,
+      "loss": 0.7869,
+      "step": 7420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.632224254599666,
+      "learning_rate": 2.9897586706024763e-06,
+      "loss": 0.5967,
+      "step": 7421
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7319331472604649,
+      "learning_rate": 2.987409651230746e-06,
+      "loss": 0.6773,
+      "step": 7422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6792342451955644,
+      "learning_rate": 2.9850613929660366e-06,
+      "loss": 0.698,
+      "step": 7423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7380245477625635,
+      "learning_rate": 2.9827138960632076e-06,
+      "loss": 0.6786,
+      "step": 7424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9112873659596736,
+      "learning_rate": 2.980367160777048e-06,
+      "loss": 0.8075,
+      "step": 7425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4329946044809587,
+      "learning_rate": 2.978021187362261e-06,
+      "loss": 0.5534,
+      "step": 7426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4934816434283067,
+      "learning_rate": 2.97567597607346e-06,
+      "loss": 0.5699,
+      "step": 7427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6117630516190384,
+      "learning_rate": 2.9733315271651863e-06,
+      "loss": 0.6782,
+      "step": 7428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8126687606012002,
+      "learning_rate": 2.9709878408918867e-06,
+      "loss": 0.7555,
+      "step": 7429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6912605643295366,
+      "learning_rate": 2.968644917507937e-06,
+      "loss": 0.746,
+      "step": 7430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.626182602352006,
+      "learning_rate": 2.966302757267625e-06,
+      "loss": 0.661,
+      "step": 7431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5037473745802517,
+      "learning_rate": 2.963961360425153e-06,
+      "loss": 0.8026,
+      "step": 7432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.559354351532829,
+      "learning_rate": 2.961620727234645e-06,
+      "loss": 0.725,
+      "step": 7433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9363196714314468,
+      "learning_rate": 2.9592808579501364e-06,
+      "loss": 0.7002,
+      "step": 7434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5776840241249293,
+      "learning_rate": 2.956941752825587e-06,
+      "loss": 0.634,
+      "step": 7435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7381703510621536,
+      "learning_rate": 2.9546034121148714e-06,
+      "loss": 0.7127,
+      "step": 7436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7190856091505584,
+      "learning_rate": 2.952265836071776e-06,
+      "loss": 0.6978,
+      "step": 7437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6735200753638035,
+      "learning_rate": 2.9499290249500078e-06,
+      "loss": 0.6708,
+      "step": 7438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7328932677954882,
+      "learning_rate": 2.9475929790031975e-06,
+      "loss": 0.8166,
+      "step": 7439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5408188030507148,
+      "learning_rate": 2.945257698484879e-06,
+      "loss": 0.7174,
+      "step": 7440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5689412370963598,
+      "learning_rate": 2.9429231836485174e-06,
+      "loss": 0.6592,
+      "step": 7441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6627008111717447,
+      "learning_rate": 2.9405894347474793e-06,
+      "loss": 0.6254,
+      "step": 7442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.837083781585254,
+      "learning_rate": 2.938256452035062e-06,
+      "loss": 0.6579,
+      "step": 7443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7666715035355245,
+      "learning_rate": 2.9359242357644757e-06,
+      "loss": 0.7378,
+      "step": 7444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6119800098653143,
+      "learning_rate": 2.9335927861888414e-06,
+      "loss": 0.6269,
+      "step": 7445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7211340131128936,
+      "learning_rate": 2.9312621035612077e-06,
+      "loss": 0.8051,
+      "step": 7446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6180189522856365,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.7323,
+      "step": 7447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7639254750773472,
+      "learning_rate": 2.9266030401616762e-06,
+      "loss": 0.6893,
+      "step": 7448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7179222972882686,
+      "learning_rate": 2.924274659895454e-06,
+      "loss": 0.7707,
+      "step": 7449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6182082420055828,
+      "learning_rate": 2.9219470475885626e-06,
+      "loss": 0.7779,
+      "step": 7450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6851692872088693,
+      "learning_rate": 2.919620203493634e-06,
+      "loss": 0.6607,
+      "step": 7451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.640265505425813,
+      "learning_rate": 2.917294127863204e-06,
+      "loss": 0.7147,
+      "step": 7452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8049523286991427,
+      "learning_rate": 2.9149688209497353e-06,
+      "loss": 0.7451,
+      "step": 7453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7823267160412626,
+      "learning_rate": 2.9126442830056066e-06,
+      "loss": 0.6662,
+      "step": 7454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5495471833684278,
+      "learning_rate": 2.910320514283104e-06,
+      "loss": 0.7109,
+      "step": 7455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5447094856687085,
+      "learning_rate": 2.9079975150344433e-06,
+      "loss": 0.6821,
+      "step": 7456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.858363830339255,
+      "learning_rate": 2.905675285511742e-06,
+      "loss": 0.6572,
+      "step": 7457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.822158149306591,
+      "learning_rate": 2.9033538259670458e-06,
+      "loss": 0.7401,
+      "step": 7458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5929794743645305,
+      "learning_rate": 2.901033136652316e-06,
+      "loss": 0.6964,
+      "step": 7459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5170105854274156,
+      "learning_rate": 2.89871321781942e-06,
+      "loss": 0.6719,
+      "step": 7460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6390930304050815,
+      "learning_rate": 2.8963940697201532e-06,
+      "loss": 0.7041,
+      "step": 7461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7532602718217476,
+      "learning_rate": 2.894075692606225e-06,
+      "loss": 0.6553,
+      "step": 7462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7868171723470738,
+      "learning_rate": 2.891758086729253e-06,
+      "loss": 0.8173,
+      "step": 7463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.828714452933475,
+      "learning_rate": 2.8894412523407833e-06,
+      "loss": 0.7114,
+      "step": 7464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6684999336577175,
+      "learning_rate": 2.8871251896922657e-06,
+      "loss": 0.6909,
+      "step": 7465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6634060136535094,
+      "learning_rate": 2.884809899035077e-06,
+      "loss": 0.7066,
+      "step": 7466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5578753536708678,
+      "learning_rate": 2.882495380620507e-06,
+      "loss": 0.629,
+      "step": 7467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.856286344143219,
+      "learning_rate": 2.8801816346997547e-06,
+      "loss": 0.767,
+      "step": 7468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7429771399973584,
+      "learning_rate": 2.877868661523947e-06,
+      "loss": 0.7595,
+      "step": 7469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6852910997798851,
+      "learning_rate": 2.875556461344117e-06,
+      "loss": 0.6857,
+      "step": 7470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7373081083150153,
+      "learning_rate": 2.873245034411217e-06,
+      "loss": 0.6821,
+      "step": 7471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.725776716881047,
+      "learning_rate": 2.870934380976124e-06,
+      "loss": 0.6353,
+      "step": 7472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5202372783896243,
+      "learning_rate": 2.868624501289613e-06,
+      "loss": 0.6614,
+      "step": 7473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7779656779653663,
+      "learning_rate": 2.8663153956023938e-06,
+      "loss": 0.7451,
+      "step": 7474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.755161174321142,
+      "learning_rate": 2.8640070641650776e-06,
+      "loss": 0.7465,
+      "step": 7475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6513658016609747,
+      "learning_rate": 2.8616995072281996e-06,
+      "loss": 0.7849,
+      "step": 7476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8120431805519797,
+      "learning_rate": 2.8593927250422126e-06,
+      "loss": 0.7905,
+      "step": 7477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5569497819936433,
+      "learning_rate": 2.8570867178574757e-06,
+      "loss": 0.7119,
+      "step": 7478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.573670506635637,
+      "learning_rate": 2.854781485924273e-06,
+      "loss": 0.6166,
+      "step": 7479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6970547434025136,
+      "learning_rate": 2.852477029492804e-06,
+      "loss": 0.8129,
+      "step": 7480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6728848081611243,
+      "learning_rate": 2.8501733488131755e-06,
+      "loss": 0.6801,
+      "step": 7481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.786059631586086,
+      "learning_rate": 2.847870444135422e-06,
+      "loss": 0.7519,
+      "step": 7482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6756514266528544,
+      "learning_rate": 2.8455683157094827e-06,
+      "loss": 0.7224,
+      "step": 7483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5194335545928992,
+      "learning_rate": 2.8432669637852208e-06,
+      "loss": 0.5874,
+      "step": 7484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6876242980806946,
+      "learning_rate": 2.840966388612413e-06,
+      "loss": 0.7177,
+      "step": 7485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8113180493913745,
+      "learning_rate": 2.838666590440746e-06,
+      "loss": 0.7245,
+      "step": 7486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6203584425862203,
+      "learning_rate": 2.8363675695198322e-06,
+      "loss": 0.6698,
+      "step": 7487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4849723093207292,
+      "learning_rate": 2.83406932609919e-06,
+      "loss": 0.6125,
+      "step": 7488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.850867969596928,
+      "learning_rate": 2.8317718604282596e-06,
+      "loss": 0.6292,
+      "step": 7489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5714885693578153,
+      "learning_rate": 2.8294751727563986e-06,
+      "loss": 0.6172,
+      "step": 7490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.609748364086573,
+      "learning_rate": 2.82717926333287e-06,
+      "loss": 0.5915,
+      "step": 7491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7166256303963017,
+      "learning_rate": 2.824884132406862e-06,
+      "loss": 0.6597,
+      "step": 7492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9578253922637714,
+      "learning_rate": 2.8225897802274784e-06,
+      "loss": 0.8443,
+      "step": 7493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.815363824497048,
+      "learning_rate": 2.820296207043728e-06,
+      "loss": 0.7102,
+      "step": 7494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.634902000615274,
+      "learning_rate": 2.8180034131045463e-06,
+      "loss": 0.7201,
+      "step": 7495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5772494222413496,
+      "learning_rate": 2.8157113986587846e-06,
+      "loss": 0.65,
+      "step": 7496
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5834670448103767,
+      "learning_rate": 2.8134201639551974e-06,
+      "loss": 0.6062,
+      "step": 7497
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7440391226487488,
+      "learning_rate": 2.811129709242465e-06,
+      "loss": 0.6941,
+      "step": 7498
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7803969915245104,
+      "learning_rate": 2.8088400347691845e-06,
+      "loss": 0.6201,
+      "step": 7499
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.646599239304442,
+      "learning_rate": 2.8065511407838584e-06,
+      "loss": 0.6177,
+      "step": 7500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6726681217351695,
+      "learning_rate": 2.8042630275349168e-06,
+      "loss": 0.7302,
+      "step": 7501
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6567772383571295,
+      "learning_rate": 2.801975695270691e-06,
+      "loss": 0.6778,
+      "step": 7502
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7253150467123428,
+      "learning_rate": 2.7996891442394393e-06,
+      "loss": 0.7049,
+      "step": 7503
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6743500942376304,
+      "learning_rate": 2.7974033746893326e-06,
+      "loss": 0.702,
+      "step": 7504
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6545542816533727,
+      "learning_rate": 2.7951183868684516e-06,
+      "loss": 0.8051,
+      "step": 7505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5076080445687705,
+      "learning_rate": 2.792834181024798e-06,
+      "loss": 0.6293,
+      "step": 7506
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7708757256910792,
+      "learning_rate": 2.7905507574062895e-06,
+      "loss": 0.7554,
+      "step": 7507
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7475955529744212,
+      "learning_rate": 2.7882681162607494e-06,
+      "loss": 0.7865,
+      "step": 7508
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.472225771004639,
+      "learning_rate": 2.7859862578359276e-06,
+      "loss": 0.6569,
+      "step": 7509
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6686558241730758,
+      "learning_rate": 2.783705182379485e-06,
+      "loss": 0.7829,
+      "step": 7510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.695276979117581,
+      "learning_rate": 2.781424890138992e-06,
+      "loss": 0.7095,
+      "step": 7511
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.646657766974499,
+      "learning_rate": 2.779145381361942e-06,
+      "loss": 0.6725,
+      "step": 7512
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.783718666727429,
+      "learning_rate": 2.7768666562957413e-06,
+      "loss": 0.6186,
+      "step": 7513
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4484205856045251,
+      "learning_rate": 2.774588715187705e-06,
+      "loss": 0.5485,
+      "step": 7514
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5731511109970013,
+      "learning_rate": 2.7723115582850714e-06,
+      "loss": 0.7341,
+      "step": 7515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.691000533782961,
+      "learning_rate": 2.7700351858349917e-06,
+      "loss": 0.6658,
+      "step": 7516
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7362006623624997,
+      "learning_rate": 2.767759598084526e-06,
+      "loss": 0.729,
+      "step": 7517
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8996942080774435,
+      "learning_rate": 2.7654847952806597e-06,
+      "loss": 0.7411,
+      "step": 7518
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6757634471720677,
+      "learning_rate": 2.763210777670281e-06,
+      "loss": 0.7495,
+      "step": 7519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6398472456760649,
+      "learning_rate": 2.7609375455002007e-06,
+      "loss": 0.6899,
+      "step": 7520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6280316209407817,
+      "learning_rate": 2.758665099017147e-06,
+      "loss": 0.7565,
+      "step": 7521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7555060068451787,
+      "learning_rate": 2.7563934384677525e-06,
+      "loss": 0.6801,
+      "step": 7522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.503827561698193,
+      "learning_rate": 2.7541225640985757e-06,
+      "loss": 0.6543,
+      "step": 7523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3849744538887527,
+      "learning_rate": 2.7518524761560784e-06,
+      "loss": 0.6735,
+      "step": 7524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4838992879970025,
+      "learning_rate": 2.749583174886646e-06,
+      "loss": 0.6396,
+      "step": 7525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6575210361822101,
+      "learning_rate": 2.7473146605365807e-06,
+      "loss": 0.7453,
+      "step": 7526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7375430861012504,
+      "learning_rate": 2.7450469333520856e-06,
+      "loss": 0.7819,
+      "step": 7527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6032934711359028,
+      "learning_rate": 2.7427799935792943e-06,
+      "loss": 0.7154,
+      "step": 7528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7372976790322099,
+      "learning_rate": 2.7405138414642405e-06,
+      "loss": 0.6819,
+      "step": 7529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.734259457654482,
+      "learning_rate": 2.738248477252885e-06,
+      "loss": 0.762,
+      "step": 7530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5259134220595947,
+      "learning_rate": 2.7359839011910983e-06,
+      "loss": 0.6118,
+      "step": 7531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6101814510917998,
+      "learning_rate": 2.7337201135246604e-06,
+      "loss": 0.7868,
+      "step": 7532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.536141528487024,
+      "learning_rate": 2.731457114499272e-06,
+      "loss": 0.6493,
+      "step": 7533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8158612352878118,
+      "learning_rate": 2.729194904360549e-06,
+      "loss": 0.7682,
+      "step": 7534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5484988134903386,
+      "learning_rate": 2.726933483354014e-06,
+      "loss": 0.7837,
+      "step": 7535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6539243202922078,
+      "learning_rate": 2.7246728517251142e-06,
+      "loss": 0.6873,
+      "step": 7536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6073679457280263,
+      "learning_rate": 2.722413009719199e-06,
+      "loss": 0.6061,
+      "step": 7537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6858862021617782,
+      "learning_rate": 2.720153957581544e-06,
+      "loss": 0.6199,
+      "step": 7538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6222117288119169,
+      "learning_rate": 2.7178956955573367e-06,
+      "loss": 0.7623,
+      "step": 7539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.733992369513456,
+      "learning_rate": 2.715638223891668e-06,
+      "loss": 0.7118,
+      "step": 7540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4844849837150704,
+      "learning_rate": 2.7133815428295585e-06,
+      "loss": 0.6526,
+      "step": 7541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6291670828286082,
+      "learning_rate": 2.711125652615929e-06,
+      "loss": 0.7378,
+      "step": 7542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6348415045516584,
+      "learning_rate": 2.708870553495625e-06,
+      "loss": 0.7867,
+      "step": 7543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6860056402117856,
+      "learning_rate": 2.706616245713406e-06,
+      "loss": 0.6777,
+      "step": 7544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.604620396060263,
+      "learning_rate": 2.704362729513934e-06,
+      "loss": 0.712,
+      "step": 7545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4305280672456266,
+      "learning_rate": 2.7021100051417994e-06,
+      "loss": 0.7249,
+      "step": 7546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7898274507982526,
+      "learning_rate": 2.6998580728414948e-06,
+      "loss": 0.7568,
+      "step": 7547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6675353834219373,
+      "learning_rate": 2.697606932857435e-06,
+      "loss": 0.705,
+      "step": 7548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6749144420557143,
+      "learning_rate": 2.695356585433949e-06,
+      "loss": 0.8273,
+      "step": 7549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6316908542704445,
+      "learning_rate": 2.693107030815271e-06,
+      "loss": 0.6924,
+      "step": 7550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7561594210981057,
+      "learning_rate": 2.6908582692455575e-06,
+      "loss": 0.7581,
+      "step": 7551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6024697973330937,
+      "learning_rate": 2.688610300968879e-06,
+      "loss": 0.6427,
+      "step": 7552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6147055907512202,
+      "learning_rate": 2.6863631262292136e-06,
+      "loss": 0.7513,
+      "step": 7553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.652596883620534,
+      "learning_rate": 2.6841167452704597e-06,
+      "loss": 0.6232,
+      "step": 7554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.783816422018661,
+      "learning_rate": 2.681871158336423e-06,
+      "loss": 0.7052,
+      "step": 7555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7142749273013935,
+      "learning_rate": 2.6796263656708297e-06,
+      "loss": 0.7456,
+      "step": 7556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6310600719430486,
+      "learning_rate": 2.6773823675173195e-06,
+      "loss": 0.6341,
+      "step": 7557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7806935404941204,
+      "learning_rate": 2.6751391641194378e-06,
+      "loss": 0.7257,
+      "step": 7558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6827389065638336,
+      "learning_rate": 2.6728967557206544e-06,
+      "loss": 0.6987,
+      "step": 7559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6514931921887344,
+      "learning_rate": 2.670655142564341e-06,
+      "loss": 0.7026,
+      "step": 7560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.616645078206799,
+      "learning_rate": 2.6684143248937955e-06,
+      "loss": 0.7544,
+      "step": 7561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7718859583686468,
+      "learning_rate": 2.6661743029522236e-06,
+      "loss": 0.6108,
+      "step": 7562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.68104639669912,
+      "learning_rate": 2.663935076982741e-06,
+      "loss": 0.8155,
+      "step": 7563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8021958735996275,
+      "learning_rate": 2.6616966472283843e-06,
+      "loss": 0.7101,
+      "step": 7564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5721868545560538,
+      "learning_rate": 2.6594590139320962e-06,
+      "loss": 0.6331,
+      "step": 7565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.777546054232048,
+      "learning_rate": 2.6572221773367392e-06,
+      "loss": 0.7279,
+      "step": 7566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.845961818905768,
+      "learning_rate": 2.6549861376850883e-06,
+      "loss": 0.7134,
+      "step": 7567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6594238425370311,
+      "learning_rate": 2.6527508952198276e-06,
+      "loss": 0.7744,
+      "step": 7568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5296321223944114,
+      "learning_rate": 2.650516450183558e-06,
+      "loss": 0.6848,
+      "step": 7569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5650590985861443,
+      "learning_rate": 2.6482828028187977e-06,
+      "loss": 0.6022,
+      "step": 7570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7019456186669477,
+      "learning_rate": 2.6460499533679685e-06,
+      "loss": 0.6716,
+      "step": 7571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.716828661162016,
+      "learning_rate": 2.6438179020734165e-06,
+      "loss": 0.6754,
+      "step": 7572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5025504069938012,
+      "learning_rate": 2.641586649177391e-06,
+      "loss": 0.7557,
+      "step": 7573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6821351398535194,
+      "learning_rate": 2.6393561949220625e-06,
+      "loss": 0.6906,
+      "step": 7574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6593300718919421,
+      "learning_rate": 2.637126539549514e-06,
+      "loss": 0.7463,
+      "step": 7575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0347185767831557,
+      "learning_rate": 2.6348976833017336e-06,
+      "loss": 0.7112,
+      "step": 7576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7191074392544898,
+      "learning_rate": 2.6326696264206352e-06,
+      "loss": 0.8615,
+      "step": 7577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8354634059261874,
+      "learning_rate": 2.630442369148034e-06,
+      "loss": 0.7765,
+      "step": 7578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.61215000399093,
+      "learning_rate": 2.628215911725668e-06,
+      "loss": 0.7777,
+      "step": 7579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6966535322409748,
+      "learning_rate": 2.6259902543951844e-06,
+      "loss": 0.7105,
+      "step": 7580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6420719462317774,
+      "learning_rate": 2.6237653973981393e-06,
+      "loss": 0.6423,
+      "step": 7581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.729672866162764,
+      "learning_rate": 2.6215413409760113e-06,
+      "loss": 0.755,
+      "step": 7582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8109345442166387,
+      "learning_rate": 2.6193180853701825e-06,
+      "loss": 0.7652,
+      "step": 7583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.670833498715669,
+      "learning_rate": 2.617095630821955e-06,
+      "loss": 0.7259,
+      "step": 7584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.626731046412688,
+      "learning_rate": 2.6148739775725427e-06,
+      "loss": 0.6965,
+      "step": 7585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7216672219356004,
+      "learning_rate": 2.6126531258630682e-06,
+      "loss": 0.5594,
+      "step": 7586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6056511431005087,
+      "learning_rate": 2.6104330759345707e-06,
+      "loss": 0.7508,
+      "step": 7587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.727656277588795,
+      "learning_rate": 2.6082138280280068e-06,
+      "loss": 0.703,
+      "step": 7588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5854088043978658,
+      "learning_rate": 2.6059953823842333e-06,
+      "loss": 0.6803,
+      "step": 7589
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.599543277038824,
+      "learning_rate": 2.6037777392440346e-06,
+      "loss": 0.6838,
+      "step": 7590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4815510408464663,
+      "learning_rate": 2.6015608988480956e-06,
+      "loss": 0.6471,
+      "step": 7591
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.789488671669432,
+      "learning_rate": 2.599344861437023e-06,
+      "loss": 0.8288,
+      "step": 7592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7897867030103445,
+      "learning_rate": 2.5971296272513347e-06,
+      "loss": 0.6958,
+      "step": 7593
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7961417034922662,
+      "learning_rate": 2.5949151965314547e-06,
+      "loss": 0.6808,
+      "step": 7594
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4180384475284087,
+      "learning_rate": 2.5927015695177305e-06,
+      "loss": 0.6841,
+      "step": 7595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6465686467448446,
+      "learning_rate": 2.5904887464504115e-06,
+      "loss": 0.6897,
+      "step": 7596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6561505477920748,
+      "learning_rate": 2.5882767275696663e-06,
+      "loss": 0.7551,
+      "step": 7597
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6116171921167224,
+      "learning_rate": 2.5860655131155788e-06,
+      "loss": 0.6677,
+      "step": 7598
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7015043649888153,
+      "learning_rate": 2.5838551033281366e-06,
+      "loss": 0.7358,
+      "step": 7599
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6427765139219823,
+      "learning_rate": 2.581645498447247e-06,
+      "loss": 0.7323,
+      "step": 7600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6021299536685927,
+      "learning_rate": 2.579436698712732e-06,
+      "loss": 0.7018,
+      "step": 7601
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.544168120396529,
+      "learning_rate": 2.577228704364314e-06,
+      "loss": 0.7018,
+      "step": 7602
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5393286785916402,
+      "learning_rate": 2.575021515641646e-06,
+      "loss": 0.6846,
+      "step": 7603
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5298286568053745,
+      "learning_rate": 2.572815132784274e-06,
+      "loss": 0.56,
+      "step": 7604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.846411894237571,
+      "learning_rate": 2.570609556031672e-06,
+      "loss": 0.7396,
+      "step": 7605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7371574415157183,
+      "learning_rate": 2.568404785623224e-06,
+      "loss": 0.7063,
+      "step": 7606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6308986377759827,
+      "learning_rate": 2.566200821798216e-06,
+      "loss": 0.6815,
+      "step": 7607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6082901733996213,
+      "learning_rate": 2.5639976647958564e-06,
+      "loss": 0.6715,
+      "step": 7608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7413064688558009,
+      "learning_rate": 2.5617953148552677e-06,
+      "loss": 0.7327,
+      "step": 7609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5715591340768313,
+      "learning_rate": 2.559593772215474e-06,
+      "loss": 0.6851,
+      "step": 7610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7550216930518252,
+      "learning_rate": 2.557393037115421e-06,
+      "loss": 0.7452,
+      "step": 7611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4456545312409386,
+      "learning_rate": 2.5551931097939676e-06,
+      "loss": 0.5707,
+      "step": 7612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5996138048373714,
+      "learning_rate": 2.552993990489876e-06,
+      "loss": 0.7159,
+      "step": 7613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.526924196735615,
+      "learning_rate": 2.5507956794418275e-06,
+      "loss": 0.6277,
+      "step": 7614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6733090546228477,
+      "learning_rate": 2.548598176888419e-06,
+      "loss": 0.7264,
+      "step": 7615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5680082178643975,
+      "learning_rate": 2.5464014830681496e-06,
+      "loss": 0.7707,
+      "step": 7616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7232490509422664,
+      "learning_rate": 2.544205598219437e-06,
+      "loss": 0.7231,
+      "step": 7617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7050713253171732,
+      "learning_rate": 2.5420105225806134e-06,
+      "loss": 0.6982,
+      "step": 7618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6466518129271135,
+      "learning_rate": 2.5398162563899154e-06,
+      "loss": 0.7152,
+      "step": 7619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.640540070141524,
+      "learning_rate": 2.5376227998854987e-06,
+      "loss": 0.6974,
+      "step": 7620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6753118861849827,
+      "learning_rate": 2.5354301533054315e-06,
+      "loss": 0.7223,
+      "step": 7621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6313335673420037,
+      "learning_rate": 2.5332383168876853e-06,
+      "loss": 0.6825,
+      "step": 7622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.856265822522839,
+      "learning_rate": 2.5310472908701555e-06,
+      "loss": 0.7676,
+      "step": 7623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5648429901605199,
+      "learning_rate": 2.5288570754906374e-06,
+      "loss": 0.7181,
+      "step": 7624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5724235897268772,
+      "learning_rate": 2.5266676709868486e-06,
+      "loss": 0.697,
+      "step": 7625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6165299290103299,
+      "learning_rate": 2.524479077596418e-06,
+      "loss": 0.7445,
+      "step": 7626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7399987056517636,
+      "learning_rate": 2.5222912955568757e-06,
+      "loss": 0.7388,
+      "step": 7627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5705171211115758,
+      "learning_rate": 2.5201043251056756e-06,
+      "loss": 0.7054,
+      "step": 7628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7790089942542553,
+      "learning_rate": 2.517918166480181e-06,
+      "loss": 0.7001,
+      "step": 7629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6494147915260875,
+      "learning_rate": 2.51573281991766e-06,
+      "loss": 0.7247,
+      "step": 7630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5286186025147486,
+      "learning_rate": 2.5135482856553027e-06,
+      "loss": 0.6604,
+      "step": 7631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7716342052578273,
+      "learning_rate": 2.511364563930202e-06,
+      "loss": 0.7095,
+      "step": 7632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5227796689934394,
+      "learning_rate": 2.509181654979368e-06,
+      "loss": 0.6938,
+      "step": 7633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7793672610158484,
+      "learning_rate": 2.5069995590397255e-06,
+      "loss": 0.7239,
+      "step": 7634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7725492488200256,
+      "learning_rate": 2.504818276348101e-06,
+      "loss": 0.6739,
+      "step": 7635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6668210564804644,
+      "learning_rate": 2.5026378071412428e-06,
+      "loss": 0.653,
+      "step": 7636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.543062332712173,
+      "learning_rate": 2.5004581516558033e-06,
+      "loss": 0.509,
+      "step": 7637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.767271637712074,
+      "learning_rate": 2.498279310128351e-06,
+      "loss": 0.6499,
+      "step": 7638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6216565293780412,
+      "learning_rate": 2.496101282795369e-06,
+      "loss": 0.7518,
+      "step": 7639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.85997349227327,
+      "learning_rate": 2.493924069893241e-06,
+      "loss": 0.6976,
+      "step": 7640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.661397393733398,
+      "learning_rate": 2.4917476716582776e-06,
+      "loss": 0.7266,
+      "step": 7641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5895203634704989,
+      "learning_rate": 2.4895720883266848e-06,
+      "loss": 0.6325,
+      "step": 7642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.680372868464412,
+      "learning_rate": 2.4873973201345924e-06,
+      "loss": 0.7073,
+      "step": 7643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6119049547484752,
+      "learning_rate": 2.4852233673180394e-06,
+      "loss": 0.5843,
+      "step": 7644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6759906839769962,
+      "learning_rate": 2.483050230112969e-06,
+      "loss": 0.7405,
+      "step": 7645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8364949521475253,
+      "learning_rate": 2.480877908755245e-06,
+      "loss": 0.72,
+      "step": 7646
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6927374995881244,
+      "learning_rate": 2.4787064034806397e-06,
+      "loss": 0.779,
+      "step": 7647
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5775482181187785,
+      "learning_rate": 2.4765357145248327e-06,
+      "loss": 0.6775,
+      "step": 7648
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9862724855522034,
+      "learning_rate": 2.474365842123422e-06,
+      "loss": 0.7681,
+      "step": 7649
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5724179992615193,
+      "learning_rate": 2.472196786511909e-06,
+      "loss": 0.6553,
+      "step": 7650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.711717378165331,
+      "learning_rate": 2.4700285479257125e-06,
+      "loss": 0.6962,
+      "step": 7651
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6741527308645596,
+      "learning_rate": 2.467861126600165e-06,
+      "loss": 0.6617,
+      "step": 7652
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7115676529691866,
+      "learning_rate": 2.4656945227705e-06,
+      "loss": 0.7768,
+      "step": 7653
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.575870000219261,
+      "learning_rate": 2.463528736671874e-06,
+      "loss": 0.6564,
+      "step": 7654
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6279772119924147,
+      "learning_rate": 2.4613637685393433e-06,
+      "loss": 0.704,
+      "step": 7655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.601742728230388,
+      "learning_rate": 2.459199618607885e-06,
+      "loss": 0.6312,
+      "step": 7656
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6462873763439139,
+      "learning_rate": 2.4570362871123856e-06,
+      "loss": 0.745,
+      "step": 7657
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7267682492023038,
+      "learning_rate": 2.454873774287635e-06,
+      "loss": 0.7738,
+      "step": 7658
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8839223215834573,
+      "learning_rate": 2.452712080368348e-06,
+      "loss": 0.7788,
+      "step": 7659
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6246102550786494,
+      "learning_rate": 2.4505512055891345e-06,
+      "loss": 0.7169,
+      "step": 7660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.586426427748253,
+      "learning_rate": 2.448391150184527e-06,
+      "loss": 0.7218,
+      "step": 7661
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6988982131523829,
+      "learning_rate": 2.44623191438897e-06,
+      "loss": 0.6827,
+      "step": 7662
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.665934565217917,
+      "learning_rate": 2.4440734984368075e-06,
+      "loss": 0.596,
+      "step": 7663
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.603377046407116,
+      "learning_rate": 2.441915902562305e-06,
+      "loss": 0.6256,
+      "step": 7664
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5480250415462409,
+      "learning_rate": 2.439759126999639e-06,
+      "loss": 0.6626,
+      "step": 7665
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5877128801469451,
+      "learning_rate": 2.4376031719828885e-06,
+      "loss": 0.5916,
+      "step": 7666
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.631419448675794,
+      "learning_rate": 2.4354480377460544e-06,
+      "loss": 0.665,
+      "step": 7667
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5382330429891682,
+      "learning_rate": 2.433293724523036e-06,
+      "loss": 0.6861,
+      "step": 7668
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7312481669876751,
+      "learning_rate": 2.431140232547653e-06,
+      "loss": 0.7801,
+      "step": 7669
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9421959679511438,
+      "learning_rate": 2.4289875620536375e-06,
+      "loss": 0.7,
+      "step": 7670
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.691814528916687,
+      "learning_rate": 2.426835713274622e-06,
+      "loss": 0.6839,
+      "step": 7671
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5587877855016785,
+      "learning_rate": 2.424684686444162e-06,
+      "loss": 0.6143,
+      "step": 7672
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6222870119900274,
+      "learning_rate": 2.422534481795711e-06,
+      "loss": 0.6179,
+      "step": 7673
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6791909756698284,
+      "learning_rate": 2.420385099562644e-06,
+      "loss": 0.6535,
+      "step": 7674
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6326619615601652,
+      "learning_rate": 2.4182365399782448e-06,
+      "loss": 0.7029,
+      "step": 7675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5625829982073207,
+      "learning_rate": 2.4160888032757014e-06,
+      "loss": 0.7023,
+      "step": 7676
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5667168902829505,
+      "learning_rate": 2.413941889688123e-06,
+      "loss": 0.7171,
+      "step": 7677
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6014556397474373,
+      "learning_rate": 2.4117957994485154e-06,
+      "loss": 0.692,
+      "step": 7678
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8147424714482825,
+      "learning_rate": 2.409650532789808e-06,
+      "loss": 0.7188,
+      "step": 7679
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.645615433366562,
+      "learning_rate": 2.4075060899448388e-06,
+      "loss": 0.6687,
+      "step": 7680
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6718486820675713,
+      "learning_rate": 2.4053624711463476e-06,
+      "loss": 0.6324,
+      "step": 7681
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8500146601648426,
+      "learning_rate": 2.4032196766269924e-06,
+      "loss": 0.7548,
+      "step": 7682
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.580356225872168,
+      "learning_rate": 2.4010777066193437e-06,
+      "loss": 0.7034,
+      "step": 7683
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5901655451727288,
+      "learning_rate": 2.3989365613558745e-06,
+      "loss": 0.8001,
+      "step": 7684
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6629579059500412,
+      "learning_rate": 2.396796241068976e-06,
+      "loss": 0.696,
+      "step": 7685
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.822392458462985,
+      "learning_rate": 2.394656745990942e-06,
+      "loss": 0.7652,
+      "step": 7686
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.62098466809815,
+      "learning_rate": 2.3925180763539847e-06,
+      "loss": 0.6656,
+      "step": 7687
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4926140404062171,
+      "learning_rate": 2.3903802323902246e-06,
+      "loss": 0.6894,
+      "step": 7688
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5855805740664932,
+      "learning_rate": 2.388243214331686e-06,
+      "loss": 0.6984,
+      "step": 7689
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6375661414055525,
+      "learning_rate": 2.3861070224103156e-06,
+      "loss": 0.5445,
+      "step": 7690
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.593248155857221,
+      "learning_rate": 2.3839716568579563e-06,
+      "loss": 0.6813,
+      "step": 7691
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6421393986690094,
+      "learning_rate": 2.3818371179063728e-06,
+      "loss": 0.6451,
+      "step": 7692
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7259233059031243,
+      "learning_rate": 2.379703405787238e-06,
+      "loss": 0.6793,
+      "step": 7693
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7278740003752728,
+      "learning_rate": 2.3775705207321275e-06,
+      "loss": 0.6303,
+      "step": 7694
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6994493656161245,
+      "learning_rate": 2.3754384629725392e-06,
+      "loss": 0.7465,
+      "step": 7695
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3677942954050732,
+      "learning_rate": 2.373307232739868e-06,
+      "loss": 0.6011,
+      "step": 7696
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5774428491429453,
+      "learning_rate": 2.371176830265427e-06,
+      "loss": 0.6296,
+      "step": 7697
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6453817635216235,
+      "learning_rate": 2.3690472557804436e-06,
+      "loss": 0.6988,
+      "step": 7698
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4742886891745193,
+      "learning_rate": 2.366918509516043e-06,
+      "loss": 0.676,
+      "step": 7699
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7364301584235993,
+      "learning_rate": 2.36479059170327e-06,
+      "loss": 0.6986,
+      "step": 7700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6897643385036853,
+      "learning_rate": 2.3626635025730803e-06,
+      "loss": 0.7172,
+      "step": 7701
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7457534652300224,
+      "learning_rate": 2.360537242356329e-06,
+      "loss": 0.6111,
+      "step": 7702
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7300879418703208,
+      "learning_rate": 2.358411811283796e-06,
+      "loss": 0.8058,
+      "step": 7703
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6630483822750175,
+      "learning_rate": 2.356287209586157e-06,
+      "loss": 0.6579,
+      "step": 7704
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7146054640600066,
+      "learning_rate": 2.354163437494006e-06,
+      "loss": 0.6689,
+      "step": 7705
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7399670370764073,
+      "learning_rate": 2.3520404952378495e-06,
+      "loss": 0.7783,
+      "step": 7706
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.82897046288453,
+      "learning_rate": 2.349918383048094e-06,
+      "loss": 0.6876,
+      "step": 7707
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.55184455103855,
+      "learning_rate": 2.347797101155067e-06,
+      "loss": 0.635,
+      "step": 7708
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6004898009171504,
+      "learning_rate": 2.345676649788995e-06,
+      "loss": 0.6636,
+      "step": 7709
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6645234770267519,
+      "learning_rate": 2.3435570291800214e-06,
+      "loss": 0.6275,
+      "step": 7710
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6193108195328993,
+      "learning_rate": 2.3414382395582024e-06,
+      "loss": 0.6788,
+      "step": 7711
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6830045676431642,
+      "learning_rate": 2.339320281153494e-06,
+      "loss": 0.6558,
+      "step": 7712
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5169238635880982,
+      "learning_rate": 2.3372031541957684e-06,
+      "loss": 0.6622,
+      "step": 7713
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6380575150031764,
+      "learning_rate": 2.335086858914811e-06,
+      "loss": 0.7452,
+      "step": 7714
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.834230091319366,
+      "learning_rate": 2.3329713955403067e-06,
+      "loss": 0.6683,
+      "step": 7715
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.585914593940905,
+      "learning_rate": 2.3308567643018574e-06,
+      "loss": 0.6325,
+      "step": 7716
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5950944991975575,
+      "learning_rate": 2.3287429654289774e-06,
+      "loss": 0.7191,
+      "step": 7717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7535720455581632,
+      "learning_rate": 2.326629999151081e-06,
+      "loss": 0.6727,
+      "step": 7718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.650372784385992,
+      "learning_rate": 2.324517865697501e-06,
+      "loss": 0.651,
+      "step": 7719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6714604758276976,
+      "learning_rate": 2.322406565297477e-06,
+      "loss": 0.6351,
+      "step": 7720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8328149534027336,
+      "learning_rate": 2.320296098180154e-06,
+      "loss": 0.6832,
+      "step": 7721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6831044613699842,
+      "learning_rate": 2.318186464574592e-06,
+      "loss": 0.6869,
+      "step": 7722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7370014886232752,
+      "learning_rate": 2.316077664709763e-06,
+      "loss": 0.7249,
+      "step": 7723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5167056453499903,
+      "learning_rate": 2.3139696988145367e-06,
+      "loss": 0.6399,
+      "step": 7724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7606882595223425,
+      "learning_rate": 2.311862567117703e-06,
+      "loss": 0.6867,
+      "step": 7725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7430394028378702,
+      "learning_rate": 2.3097562698479625e-06,
+      "loss": 0.722,
+      "step": 7726
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8257985091538622,
+      "learning_rate": 2.307650807233913e-06,
+      "loss": 0.735,
+      "step": 7727
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.370233037411586,
+      "learning_rate": 2.3055461795040745e-06,
+      "loss": 0.6408,
+      "step": 7728
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9713638047994393,
+      "learning_rate": 2.3034423868868728e-06,
+      "loss": 0.7296,
+      "step": 7729
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5408962439710896,
+      "learning_rate": 2.301339429610637e-06,
+      "loss": 0.6946,
+      "step": 7730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7594833023077556,
+      "learning_rate": 2.2992373079036146e-06,
+      "loss": 0.7895,
+      "step": 7731
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8144959438221029,
+      "learning_rate": 2.2971360219939543e-06,
+      "loss": 0.7169,
+      "step": 7732
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6810027650585382,
+      "learning_rate": 2.295035572109718e-06,
+      "loss": 0.5644,
+      "step": 7733
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7103028604041395,
+      "learning_rate": 2.2929359584788814e-06,
+      "loss": 0.74,
+      "step": 7734
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.66925295250059,
+      "learning_rate": 2.2908371813293195e-06,
+      "loss": 0.6676,
+      "step": 7735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7461573353125468,
+      "learning_rate": 2.2887392408888233e-06,
+      "loss": 0.8088,
+      "step": 7736
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6376307701564305,
+      "learning_rate": 2.2866421373850943e-06,
+      "loss": 0.7629,
+      "step": 7737
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.536706107384853,
+      "learning_rate": 2.284545871045736e-06,
+      "loss": 0.669,
+      "step": 7738
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5271382033522722,
+      "learning_rate": 2.2824504420982683e-06,
+      "loss": 0.5886,
+      "step": 7739
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7006083419959541,
+      "learning_rate": 2.280355850770115e-06,
+      "loss": 0.7803,
+      "step": 7740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5722117495050347,
+      "learning_rate": 2.2782620972886116e-06,
+      "loss": 0.6991,
+      "step": 7741
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4054192798777325,
+      "learning_rate": 2.276169181881006e-06,
+      "loss": 0.6412,
+      "step": 7742
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6087736248439393,
+      "learning_rate": 2.274077104774446e-06,
+      "loss": 0.6813,
+      "step": 7743
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7157752209095043,
+      "learning_rate": 2.2719858661959994e-06,
+      "loss": 0.7376,
+      "step": 7744
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.514585889933833,
+      "learning_rate": 2.26989546637263e-06,
+      "loss": 0.6464,
+      "step": 7745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4337615493436142,
+      "learning_rate": 2.267805905531224e-06,
+      "loss": 0.6228,
+      "step": 7746
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6849361457896674,
+      "learning_rate": 2.2657171838985714e-06,
+      "loss": 0.722,
+      "step": 7747
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6944312615856458,
+      "learning_rate": 2.2636293017013646e-06,
+      "loss": 0.7142,
+      "step": 7748
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.879071950573806,
+      "learning_rate": 2.2615422591662175e-06,
+      "loss": 0.7979,
+      "step": 7749
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.614926120698719,
+      "learning_rate": 2.259456056519639e-06,
+      "loss": 0.6204,
+      "step": 7750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8507881781454831,
+      "learning_rate": 2.257370693988056e-06,
+      "loss": 0.7752,
+      "step": 7751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6303593430397691,
+      "learning_rate": 2.255286171797807e-06,
+      "loss": 0.6652,
+      "step": 7752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7243678063678334,
+      "learning_rate": 2.2532024901751273e-06,
+      "loss": 0.7451,
+      "step": 7753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6780366300873248,
+      "learning_rate": 2.251119649346173e-06,
+      "loss": 0.676,
+      "step": 7754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7377231196349392,
+      "learning_rate": 2.249037649536999e-06,
+      "loss": 0.7033,
+      "step": 7755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4944129500941656,
+      "learning_rate": 2.2469564909735773e-06,
+      "loss": 0.6111,
+      "step": 7756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6804228090544804,
+      "learning_rate": 2.244876173881787e-06,
+      "loss": 0.7282,
+      "step": 7757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5289001691878608,
+      "learning_rate": 2.2427966984874095e-06,
+      "loss": 0.669,
+      "step": 7758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6648677967922552,
+      "learning_rate": 2.240718065016141e-06,
+      "loss": 0.7155,
+      "step": 7759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5616347105803132,
+      "learning_rate": 2.2386402736935865e-06,
+      "loss": 0.6067,
+      "step": 7760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6999639438907512,
+      "learning_rate": 2.2365633247452546e-06,
+      "loss": 0.645,
+      "step": 7761
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7686905266747723,
+      "learning_rate": 2.2344872183965694e-06,
+      "loss": 0.6266,
+      "step": 7762
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.551309634097524,
+      "learning_rate": 2.232411954872855e-06,
+      "loss": 0.6743,
+      "step": 7763
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.683891809634969,
+      "learning_rate": 2.2303375343993515e-06,
+      "loss": 0.7688,
+      "step": 7764
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.585579032545016,
+      "learning_rate": 2.2282639572012075e-06,
+      "loss": 0.6695,
+      "step": 7765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6017477879438695,
+      "learning_rate": 2.226191223503472e-06,
+      "loss": 0.718,
+      "step": 7766
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5631122799217767,
+      "learning_rate": 2.2241193335311127e-06,
+      "loss": 0.6414,
+      "step": 7767
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6802047965642206,
+      "learning_rate": 2.2220482875089965e-06,
+      "loss": 0.629,
+      "step": 7768
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7959499571883017,
+      "learning_rate": 2.2199780856619045e-06,
+      "loss": 0.7789,
+      "step": 7769
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.700554129412106,
+      "learning_rate": 2.2179087282145283e-06,
+      "loss": 0.7761,
+      "step": 7770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7018819641783725,
+      "learning_rate": 2.2158402153914583e-06,
+      "loss": 0.7092,
+      "step": 7771
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.092281198826054,
+      "learning_rate": 2.2137725474172056e-06,
+      "loss": 0.7933,
+      "step": 7772
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6815014348058988,
+      "learning_rate": 2.2117057245161767e-06,
+      "loss": 0.669,
+      "step": 7773
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7578458090852305,
+      "learning_rate": 2.209639746912696e-06,
+      "loss": 0.7681,
+      "step": 7774
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6446787326998182,
+      "learning_rate": 2.2075746148309964e-06,
+      "loss": 0.6735,
+      "step": 7775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7808940087861367,
+      "learning_rate": 2.2055103284952094e-06,
+      "loss": 0.7404,
+      "step": 7776
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5939379532221074,
+      "learning_rate": 2.2034468881293845e-06,
+      "loss": 0.6936,
+      "step": 7777
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6748246487282306,
+      "learning_rate": 2.2013842939574783e-06,
+      "loss": 0.6473,
+      "step": 7778
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6786313202255883,
+      "learning_rate": 2.1993225462033465e-06,
+      "loss": 0.6747,
+      "step": 7779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.865528066751501,
+      "learning_rate": 2.197261645090767e-06,
+      "loss": 0.7479,
+      "step": 7780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5587845597683763,
+      "learning_rate": 2.195201590843412e-06,
+      "loss": 0.5715,
+      "step": 7781
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6196625061757,
+      "learning_rate": 2.1931423836848697e-06,
+      "loss": 0.6109,
+      "step": 7782
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7556785120047103,
+      "learning_rate": 2.19108402383864e-06,
+      "loss": 0.6765,
+      "step": 7783
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7597770933828298,
+      "learning_rate": 2.1890265115281185e-06,
+      "loss": 0.7867,
+      "step": 7784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7962624226990016,
+      "learning_rate": 2.186969846976623e-06,
+      "loss": 0.6741,
+      "step": 7785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5725184446544542,
+      "learning_rate": 2.1849140304073647e-06,
+      "loss": 0.7164,
+      "step": 7786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6603822362173248,
+      "learning_rate": 2.1828590620434742e-06,
+      "loss": 0.7547,
+      "step": 7787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8415056371952363,
+      "learning_rate": 2.180804942107989e-06,
+      "loss": 0.6835,
+      "step": 7788
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.716076411318612,
+      "learning_rate": 2.1787516708238454e-06,
+      "loss": 0.7467,
+      "step": 7789
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6953951498862214,
+      "learning_rate": 2.1766992484139015e-06,
+      "loss": 0.778,
+      "step": 7790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5108432572094461,
+      "learning_rate": 2.174647675100907e-06,
+      "loss": 0.7437,
+      "step": 7791
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.659610127336146,
+      "learning_rate": 2.172596951107534e-06,
+      "loss": 0.6632,
+      "step": 7792
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6463493242617568,
+      "learning_rate": 2.1705470766563573e-06,
+      "loss": 0.7365,
+      "step": 7793
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6967353791802484,
+      "learning_rate": 2.168498051969854e-06,
+      "loss": 0.6863,
+      "step": 7794
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7089615810906633,
+      "learning_rate": 2.166449877270416e-06,
+      "loss": 0.6617,
+      "step": 7795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6528730277901496,
+      "learning_rate": 2.1644025527803426e-06,
+      "loss": 0.5896,
+      "step": 7796
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7234934710601961,
+      "learning_rate": 2.1623560787218355e-06,
+      "loss": 0.6047,
+      "step": 7797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7342875782041085,
+      "learning_rate": 2.1603104553170108e-06,
+      "loss": 0.6267,
+      "step": 7798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.806632183962203,
+      "learning_rate": 2.1582656827878846e-06,
+      "loss": 0.7972,
+      "step": 7799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5230796084491267,
+      "learning_rate": 2.1562217613563885e-06,
+      "loss": 0.6782,
+      "step": 7800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5813890671558295,
+      "learning_rate": 2.1541786912443596e-06,
+      "loss": 0.6945,
+      "step": 7801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9117029170190478,
+      "learning_rate": 2.152136472673535e-06,
+      "loss": 0.656,
+      "step": 7802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7405050503451025,
+      "learning_rate": 2.1500951058655727e-06,
+      "loss": 0.7311,
+      "step": 7803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8246200827578696,
+      "learning_rate": 2.148054591042026e-06,
+      "loss": 0.7639,
+      "step": 7804
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5414344095577803,
+      "learning_rate": 2.1460149284243625e-06,
+      "loss": 0.6125,
+      "step": 7805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7695724635486085,
+      "learning_rate": 2.143976118233958e-06,
+      "loss": 0.7938,
+      "step": 7806
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6712057945183567,
+      "learning_rate": 2.1419381606920887e-06,
+      "loss": 0.6595,
+      "step": 7807
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7053677147235984,
+      "learning_rate": 2.1399010560199494e-06,
+      "loss": 0.749,
+      "step": 7808
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.588553082819756,
+      "learning_rate": 2.1378648044386285e-06,
+      "loss": 0.6854,
+      "step": 7809
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6490243239056168,
+      "learning_rate": 2.135829406169133e-06,
+      "loss": 0.7192,
+      "step": 7810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.829693884474,
+      "learning_rate": 2.133794861432378e-06,
+      "loss": 0.7406,
+      "step": 7811
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7410750894386053,
+      "learning_rate": 2.1317611704491727e-06,
+      "loss": 0.6912,
+      "step": 7812
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6367704761381865,
+      "learning_rate": 2.1297283334402474e-06,
+      "loss": 0.7254,
+      "step": 7813
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.62040426586708,
+      "learning_rate": 2.127696350626236e-06,
+      "loss": 0.7247,
+      "step": 7814
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6298201574818223,
+      "learning_rate": 2.1256652222276753e-06,
+      "loss": 0.6358,
+      "step": 7815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.701264821377545,
+      "learning_rate": 2.1236349484650164e-06,
+      "loss": 0.7002,
+      "step": 7816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.687077714298927,
+      "learning_rate": 2.121605529558608e-06,
+      "loss": 0.6635,
+      "step": 7817
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.522822106403037,
+      "learning_rate": 2.1195769657287145e-06,
+      "loss": 0.5933,
+      "step": 7818
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8707382783379618,
+      "learning_rate": 2.117549257195509e-06,
+      "loss": 0.7004,
+      "step": 7819
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7169068537145045,
+      "learning_rate": 2.1155224041790614e-06,
+      "loss": 0.6212,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7913526157090918,
+      "learning_rate": 2.1134964068993568e-06,
+      "loss": 0.6165,
+      "step": 7821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.68792325925831,
+      "learning_rate": 2.1114712655762893e-06,
+      "loss": 0.7724,
+      "step": 7822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6619920318546355,
+      "learning_rate": 2.10944698042965e-06,
+      "loss": 0.7583,
+      "step": 7823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5725040329920634,
+      "learning_rate": 2.1074235516791475e-06,
+      "loss": 0.6454,
+      "step": 7824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.776077694659041,
+      "learning_rate": 2.105400979544394e-06,
+      "loss": 0.6861,
+      "step": 7825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6810506472695297,
+      "learning_rate": 2.1033792642449037e-06,
+      "loss": 0.7583,
+      "step": 7826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6039372665750604,
+      "learning_rate": 2.101358406000106e-06,
+      "loss": 0.6526,
+      "step": 7827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5976535524612527,
+      "learning_rate": 2.0993384050293343e-06,
+      "loss": 0.714,
+      "step": 7828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8511945629783728,
+      "learning_rate": 2.0973192615518234e-06,
+      "loss": 0.6959,
+      "step": 7829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.709399661423479,
+      "learning_rate": 2.095300975786723e-06,
+      "loss": 0.6782,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7332709455559026,
+      "learning_rate": 2.093283547953088e-06,
+      "loss": 0.6648,
+      "step": 7831
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6192736018667757,
+      "learning_rate": 2.0912669782698737e-06,
+      "loss": 0.6277,
+      "step": 7832
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6947247635348386,
+      "learning_rate": 2.0892512669559505e-06,
+      "loss": 0.7146,
+      "step": 7833
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.654417397999407,
+      "learning_rate": 2.087236414230096e-06,
+      "loss": 0.7693,
+      "step": 7834
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6487583791677305,
+      "learning_rate": 2.0852224203109826e-06,
+      "loss": 0.6617,
+      "step": 7835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.556761764582044,
+      "learning_rate": 2.083209285417206e-06,
+      "loss": 0.7113,
+      "step": 7836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7912806691152143,
+      "learning_rate": 2.0811970097672527e-06,
+      "loss": 0.6975,
+      "step": 7837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.717649994233873,
+      "learning_rate": 2.079185593579529e-06,
+      "loss": 0.7063,
+      "step": 7838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5437100663665961,
+      "learning_rate": 2.0771750370723444e-06,
+      "loss": 0.6612,
+      "step": 7839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7449689894327503,
+      "learning_rate": 2.0751653404639073e-06,
+      "loss": 0.6719,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5018868498932967,
+      "learning_rate": 2.073156503972341e-06,
+      "loss": 0.6652,
+      "step": 7841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.636787781130712,
+      "learning_rate": 2.0711485278156774e-06,
+      "loss": 0.6727,
+      "step": 7842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6536033634879708,
+      "learning_rate": 2.069141412211846e-06,
+      "loss": 0.7434,
+      "step": 7843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6512144230931152,
+      "learning_rate": 2.067135157378691e-06,
+      "loss": 0.7674,
+      "step": 7844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5484002894263678,
+      "learning_rate": 2.065129763533956e-06,
+      "loss": 0.7699,
+      "step": 7845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5211767111624397,
+      "learning_rate": 2.0631252308952986e-06,
+      "loss": 0.6553,
+      "step": 7846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6673412897745399,
+      "learning_rate": 2.06112155968028e-06,
+      "loss": 0.6335,
+      "step": 7847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6113060699168706,
+      "learning_rate": 2.059118750106365e-06,
+      "loss": 0.6341,
+      "step": 7848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6460704313269863,
+      "learning_rate": 2.0571168023909272e-06,
+      "loss": 0.6317,
+      "step": 7849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7483784351965763,
+      "learning_rate": 2.0551157167512503e-06,
+      "loss": 0.7829,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8651438770398827,
+      "learning_rate": 2.053115493404515e-06,
+      "loss": 0.7116,
+      "step": 7851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7193660661457268,
+      "learning_rate": 2.0511161325678206e-06,
+      "loss": 0.6865,
+      "step": 7852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.622650547540691,
+      "learning_rate": 2.0491176344581608e-06,
+      "loss": 0.6429,
+      "step": 7853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.524886595091858,
+      "learning_rate": 2.047119999292444e-06,
+      "loss": 0.6307,
+      "step": 7854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6940262251806701,
+      "learning_rate": 2.0451232272874845e-06,
+      "loss": 0.6906,
+      "step": 7855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5474554934700888,
+      "learning_rate": 2.0431273186599964e-06,
+      "loss": 0.5805,
+      "step": 7856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5127942854920668,
+      "learning_rate": 2.041132273626608e-06,
+      "loss": 0.6684,
+      "step": 7857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5997306154029722,
+      "learning_rate": 2.039138092403846e-06,
+      "loss": 0.6363,
+      "step": 7858
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.875216592490199,
+      "learning_rate": 2.037144775208151e-06,
+      "loss": 0.5804,
+      "step": 7859
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6753429884560422,
+      "learning_rate": 2.035152322255868e-06,
+      "loss": 0.6385,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7548658352691162,
+      "learning_rate": 2.0331607337632407e-06,
+      "loss": 0.6925,
+      "step": 7861
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5398454208128212,
+      "learning_rate": 2.0311700099464315e-06,
+      "loss": 0.6483,
+      "step": 7862
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6031866344789765,
+      "learning_rate": 2.029180151021496e-06,
+      "loss": 0.6919,
+      "step": 7863
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5614753142650077,
+      "learning_rate": 2.027191157204406e-06,
+      "loss": 0.7527,
+      "step": 7864
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.835885954488514,
+      "learning_rate": 2.025203028711038e-06,
+      "loss": 0.7981,
+      "step": 7865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6885653035275787,
+      "learning_rate": 2.023215765757166e-06,
+      "loss": 0.7251,
+      "step": 7866
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8891035652301391,
+      "learning_rate": 2.0212293685584794e-06,
+      "loss": 0.7153,
+      "step": 7867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.974342714359956,
+      "learning_rate": 2.0192438373305747e-06,
+      "loss": 0.7538,
+      "step": 7868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7252816726702502,
+      "learning_rate": 2.0172591722889423e-06,
+      "loss": 0.7309,
+      "step": 7869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.791791144501412,
+      "learning_rate": 2.015275373648994e-06,
+      "loss": 0.7317,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.853562934767157,
+      "learning_rate": 2.0132924416260347e-06,
+      "loss": 0.8221,
+      "step": 7871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7272393998078934,
+      "learning_rate": 2.0113103764352814e-06,
+      "loss": 0.7894,
+      "step": 7872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6633087363346242,
+      "learning_rate": 2.009329178291861e-06,
+      "loss": 0.768,
+      "step": 7873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9128303825762558,
+      "learning_rate": 2.0073488474107962e-06,
+      "loss": 0.6392,
+      "step": 7874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7616167689658218,
+      "learning_rate": 2.0053693840070242e-06,
+      "loss": 0.6602,
+      "step": 7875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.699303749633239,
+      "learning_rate": 2.0033907882953818e-06,
+      "loss": 0.7439,
+      "step": 7876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6070346725208509,
+      "learning_rate": 2.0014130604906167e-06,
+      "loss": 0.7094,
+      "step": 7877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.741448073560214,
+      "learning_rate": 1.999436200807382e-06,
+      "loss": 0.7499,
+      "step": 7878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6944355963588895,
+      "learning_rate": 1.99746020946023e-06,
+      "loss": 0.7529,
+      "step": 7879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8895579187524152,
+      "learning_rate": 1.99548508666363e-06,
+      "loss": 0.7048,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9920721222550972,
+      "learning_rate": 1.993510832631944e-06,
+      "loss": 0.7369,
+      "step": 7881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5922170424114257,
+      "learning_rate": 1.9915374475794492e-06,
+      "loss": 0.6501,
+      "step": 7882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5634638693807896,
+      "learning_rate": 1.989564931720329e-06,
+      "loss": 0.6196,
+      "step": 7883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5635383047081672,
+      "learning_rate": 1.987593285268664e-06,
+      "loss": 0.7479,
+      "step": 7884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4883323785564222,
+      "learning_rate": 1.9856225084384484e-06,
+      "loss": 0.7076,
+      "step": 7885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6282107550188445,
+      "learning_rate": 1.983652601443581e-06,
+      "loss": 0.6514,
+      "step": 7886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.527130392726828,
+      "learning_rate": 1.981683564497858e-06,
+      "loss": 0.683,
+      "step": 7887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5838502653009983,
+      "learning_rate": 1.979715397814996e-06,
+      "loss": 0.737,
+      "step": 7888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6495537665659166,
+      "learning_rate": 1.977748101608601e-06,
+      "loss": 0.6973,
+      "step": 7889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7194286106142869,
+      "learning_rate": 1.9757816760921955e-06,
+      "loss": 0.7559,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5641719812122743,
+      "learning_rate": 1.973816121479207e-06,
+      "loss": 0.6447,
+      "step": 7891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5147137650724953,
+      "learning_rate": 1.971851437982961e-06,
+      "loss": 0.5577,
+      "step": 7892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7297562578736958,
+      "learning_rate": 1.9698876258166968e-06,
+      "loss": 0.698,
+      "step": 7893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.872273742313337,
+      "learning_rate": 1.967924685193552e-06,
+      "loss": 0.7417,
+      "step": 7894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5730148472491021,
+      "learning_rate": 1.965962616326574e-06,
+      "loss": 0.6882,
+      "step": 7895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7367694504986644,
+      "learning_rate": 1.9640014194287206e-06,
+      "loss": 0.7052,
+      "step": 7896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.794095966158344,
+      "learning_rate": 1.96204109471284e-06,
+      "loss": 0.6673,
+      "step": 7897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6628464531180407,
+      "learning_rate": 1.9600816423917013e-06,
+      "loss": 0.7936,
+      "step": 7898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7052997907461107,
+      "learning_rate": 1.958123062677968e-06,
+      "loss": 0.7681,
+      "step": 7899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6754284096878078,
+      "learning_rate": 1.9561653557842142e-06,
+      "loss": 0.7549,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5443046736290091,
+      "learning_rate": 1.9542085219229235e-06,
+      "loss": 0.682,
+      "step": 7901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7456115388925508,
+      "learning_rate": 1.9522525613064723e-06,
+      "loss": 0.6539,
+      "step": 7902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7079663646484742,
+      "learning_rate": 1.950297474147156e-06,
+      "loss": 0.6013,
+      "step": 7903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6242157401336126,
+      "learning_rate": 1.9483432606571627e-06,
+      "loss": 0.6876,
+      "step": 7904
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6509617781874775,
+      "learning_rate": 1.946389921048595e-06,
+      "loss": 0.6657,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5558128791049708,
+      "learning_rate": 1.944437455533459e-06,
+      "loss": 0.6951,
+      "step": 7906
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5442432225603513,
+      "learning_rate": 1.9424858643236598e-06,
+      "loss": 0.6857,
+      "step": 7907
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8953492852649612,
+      "learning_rate": 1.9405351476310154e-06,
+      "loss": 0.6654,
+      "step": 7908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6694936693614468,
+      "learning_rate": 1.9385853056672467e-06,
+      "loss": 0.6512,
+      "step": 7909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5791933585290685,
+      "learning_rate": 1.9366363386439747e-06,
+      "loss": 0.7247,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7048568831881048,
+      "learning_rate": 1.9346882467727323e-06,
+      "loss": 0.7314,
+      "step": 7911
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6997667202698497,
+      "learning_rate": 1.932741030264952e-06,
+      "loss": 0.7043,
+      "step": 7912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6307421070504355,
+      "learning_rate": 1.930794689331975e-06,
+      "loss": 0.6322,
+      "step": 7913
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0600411136366907,
+      "learning_rate": 1.9288492241850486e-06,
+      "loss": 0.8343,
+      "step": 7914
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6263978112754838,
+      "learning_rate": 1.9269046350353184e-06,
+      "loss": 0.5917,
+      "step": 7915
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7212214231886043,
+      "learning_rate": 1.9249609220938425e-06,
+      "loss": 0.6988,
+      "step": 7916
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7923502452489541,
+      "learning_rate": 1.9230180855715765e-06,
+      "loss": 0.7721,
+      "step": 7917
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6510583005199575,
+      "learning_rate": 1.9210761256793876e-06,
+      "loss": 0.6751,
+      "step": 7918
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6965142454396533,
+      "learning_rate": 1.9191350426280476e-06,
+      "loss": 0.7389,
+      "step": 7919
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5883857332917075,
+      "learning_rate": 1.9171948366282256e-06,
+      "loss": 0.6797,
+      "step": 7920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7740325510562096,
+      "learning_rate": 1.9152555078905054e-06,
+      "loss": 0.6812,
+      "step": 7921
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7902049614453566,
+      "learning_rate": 1.9133170566253665e-06,
+      "loss": 0.6932,
+      "step": 7922
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7610525327182251,
+      "learning_rate": 1.9113794830431998e-06,
+      "loss": 0.7173,
+      "step": 7923
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.737127894919951,
+      "learning_rate": 1.9094427873542997e-06,
+      "loss": 0.7971,
+      "step": 7924
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.842918131941055,
+      "learning_rate": 1.907506969768862e-06,
+      "loss": 0.6717,
+      "step": 7925
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8822392409548165,
+      "learning_rate": 1.9055720304969894e-06,
+      "loss": 0.757,
+      "step": 7926
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.72327856763112,
+      "learning_rate": 1.903637969748693e-06,
+      "loss": 0.5784,
+      "step": 7927
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5513018101818548,
+      "learning_rate": 1.901704787733879e-06,
+      "loss": 0.6345,
+      "step": 7928
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6058853335363181,
+      "learning_rate": 1.8997724846623666e-06,
+      "loss": 0.6129,
+      "step": 7929
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.732733524266622,
+      "learning_rate": 1.8978410607438812e-06,
+      "loss": 0.6901,
+      "step": 7930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7278955082537923,
+      "learning_rate": 1.895910516188042e-06,
+      "loss": 0.7151,
+      "step": 7931
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7211882101735263,
+      "learning_rate": 1.8939808512043822e-06,
+      "loss": 0.7134,
+      "step": 7932
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6376433863918067,
+      "learning_rate": 1.8920520660023412e-06,
+      "loss": 0.6377,
+      "step": 7933
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9063083868432706,
+      "learning_rate": 1.8901241607912502e-06,
+      "loss": 0.6743,
+      "step": 7934
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6773780116605972,
+      "learning_rate": 1.8881971357803575e-06,
+      "loss": 0.6674,
+      "step": 7935
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.756703253587125,
+      "learning_rate": 1.8862709911788145e-06,
+      "loss": 0.7528,
+      "step": 7936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8278159313895643,
+      "learning_rate": 1.8843457271956679e-06,
+      "loss": 0.6927,
+      "step": 7937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5222474436371374,
+      "learning_rate": 1.8824213440398776e-06,
+      "loss": 0.6128,
+      "step": 7938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5055087749630518,
+      "learning_rate": 1.880497841920308e-06,
+      "loss": 0.635,
+      "step": 7939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.817582363757226,
+      "learning_rate": 1.8785752210457208e-06,
+      "loss": 0.6813,
+      "step": 7940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.76457845246533,
+      "learning_rate": 1.8766534816247917e-06,
+      "loss": 0.6767,
+      "step": 7941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5273211143254652,
+      "learning_rate": 1.874732623866089e-06,
+      "loss": 0.6553,
+      "step": 7942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7697396621253008,
+      "learning_rate": 1.8728126479780951e-06,
+      "loss": 0.7514,
+      "step": 7943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8317315916672514,
+      "learning_rate": 1.8708935541691964e-06,
+      "loss": 0.7215,
+      "step": 7944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7793918921800997,
+      "learning_rate": 1.868975342647673e-06,
+      "loss": 0.6822,
+      "step": 7945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.813369886642259,
+      "learning_rate": 1.8670580136217232e-06,
+      "loss": 0.7838,
+      "step": 7946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6914376056588731,
+      "learning_rate": 1.865141567299442e-06,
+      "loss": 0.637,
+      "step": 7947
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8448253000887582,
+      "learning_rate": 1.8632260038888272e-06,
+      "loss": 0.7215,
+      "step": 7948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6885452785875725,
+      "learning_rate": 1.8613113235977854e-06,
+      "loss": 0.7254,
+      "step": 7949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.786429205383274,
+      "learning_rate": 1.8593975266341258e-06,
+      "loss": 0.7589,
+      "step": 7950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6108942657093994,
+      "learning_rate": 1.857484613205558e-06,
+      "loss": 0.7062,
+      "step": 7951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7621078214450279,
+      "learning_rate": 1.8555725835197024e-06,
+      "loss": 0.6255,
+      "step": 7952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7650961049535976,
+      "learning_rate": 1.8536614377840767e-06,
+      "loss": 0.6909,
+      "step": 7953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6656129367567623,
+      "learning_rate": 1.851751176206107e-06,
+      "loss": 0.6863,
+      "step": 7954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.745463279141544,
+      "learning_rate": 1.849841798993126e-06,
+      "loss": 0.8086,
+      "step": 7955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6093159687854266,
+      "learning_rate": 1.8479333063523596e-06,
+      "loss": 0.666,
+      "step": 7956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6844585878472074,
+      "learning_rate": 1.846025698490952e-06,
+      "loss": 0.7696,
+      "step": 7957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6175676462127535,
+      "learning_rate": 1.8441189756159384e-06,
+      "loss": 0.7112,
+      "step": 7958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6762261447013207,
+      "learning_rate": 1.8422131379342668e-06,
+      "loss": 0.5827,
+      "step": 7959
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6781795279750065,
+      "learning_rate": 1.840308185652787e-06,
+      "loss": 0.6993,
+      "step": 7960
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5292222257030226,
+      "learning_rate": 1.8384041189782487e-06,
+      "loss": 0.5964,
+      "step": 7961
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.781420757861504,
+      "learning_rate": 1.8365009381173104e-06,
+      "loss": 0.7575,
+      "step": 7962
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6092231268001482,
+      "learning_rate": 1.8345986432765338e-06,
+      "loss": 0.6708,
+      "step": 7963
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.569521817337266,
+      "learning_rate": 1.8326972346623806e-06,
+      "loss": 0.6602,
+      "step": 7964
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.606489507221082,
+      "learning_rate": 1.8307967124812221e-06,
+      "loss": 0.7262,
+      "step": 7965
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6885580857972373,
+      "learning_rate": 1.8288970769393267e-06,
+      "loss": 0.7442,
+      "step": 7966
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.730201721005992,
+      "learning_rate": 1.8269983282428705e-06,
+      "loss": 0.7575,
+      "step": 7967
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6215040305929973,
+      "learning_rate": 1.8251004665979378e-06,
+      "loss": 0.6663,
+      "step": 7968
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5726907533294119,
+      "learning_rate": 1.8232034922105058e-06,
+      "loss": 0.6834,
+      "step": 7969
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6229999795058794,
+      "learning_rate": 1.8213074052864654e-06,
+      "loss": 0.76,
+      "step": 7970
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5485299111513613,
+      "learning_rate": 1.8194122060316044e-06,
+      "loss": 0.6972,
+      "step": 7971
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5521491925286681,
+      "learning_rate": 1.817517894651617e-06,
+      "loss": 0.593,
+      "step": 7972
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7300321739270723,
+      "learning_rate": 1.8156244713521065e-06,
+      "loss": 0.7811,
+      "step": 7973
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.829781856253929,
+      "learning_rate": 1.8137319363385664e-06,
+      "loss": 0.6688,
+      "step": 7974
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6007145369222389,
+      "learning_rate": 1.811840289816409e-06,
+      "loss": 0.7405,
+      "step": 7975
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.597049195511601,
+      "learning_rate": 1.8099495319909377e-06,
+      "loss": 0.6841,
+      "step": 7976
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.465891069369129,
+      "learning_rate": 1.8080596630673652e-06,
+      "loss": 0.6568,
+      "step": 7977
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6293286430781235,
+      "learning_rate": 1.806170683250813e-06,
+      "loss": 0.6951,
+      "step": 7978
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6922012635236878,
+      "learning_rate": 1.8042825927462915e-06,
+      "loss": 0.7233,
+      "step": 7979
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5849598458552254,
+      "learning_rate": 1.802395391758729e-06,
+      "loss": 0.6481,
+      "step": 7980
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8039497724725537,
+      "learning_rate": 1.8005090804929525e-06,
+      "loss": 0.6401,
+      "step": 7981
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4792034062672803,
+      "learning_rate": 1.7986236591536875e-06,
+      "loss": 0.7558,
+      "step": 7982
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6653008329867707,
+      "learning_rate": 1.7967391279455715e-06,
+      "loss": 0.6519,
+      "step": 7983
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6590694279898175,
+      "learning_rate": 1.794855487073136e-06,
+      "loss": 0.7157,
+      "step": 7984
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6776772981968426,
+      "learning_rate": 1.7929727367408233e-06,
+      "loss": 0.7889,
+      "step": 7985
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6534317199844832,
+      "learning_rate": 1.79109087715298e-06,
+      "loss": 0.6789,
+      "step": 7986
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6922990817285495,
+      "learning_rate": 1.7892099085138448e-06,
+      "loss": 0.6641,
+      "step": 7987
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6048970808366634,
+      "learning_rate": 1.7873298310275755e-06,
+      "loss": 0.6557,
+      "step": 7988
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.690427387402649,
+      "learning_rate": 1.7854506448982179e-06,
+      "loss": 0.7262,
+      "step": 7989
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.691584509499384,
+      "learning_rate": 1.783572350329732e-06,
+      "loss": 0.736,
+      "step": 7990
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7445444143992106,
+      "learning_rate": 1.7816949475259793e-06,
+      "loss": 0.7128,
+      "step": 7991
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.821762724528318,
+      "learning_rate": 1.7798184366907167e-06,
+      "loss": 0.7088,
+      "step": 7992
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6726024812181441,
+      "learning_rate": 1.777942818027617e-06,
+      "loss": 0.7523,
+      "step": 7993
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6375638362892346,
+      "learning_rate": 1.776068091740244e-06,
+      "loss": 0.703,
+      "step": 7994
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7839059170132194,
+      "learning_rate": 1.7741942580320704e-06,
+      "loss": 0.7539,
+      "step": 7995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8828237716172438,
+      "learning_rate": 1.7723213171064757e-06,
+      "loss": 0.6589,
+      "step": 7996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5749266185283466,
+      "learning_rate": 1.770449269166733e-06,
+      "loss": 0.6152,
+      "step": 7997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.637547348638589,
+      "learning_rate": 1.7685781144160276e-06,
+      "loss": 0.7638,
+      "step": 7998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6348983075135142,
+      "learning_rate": 1.7667078530574432e-06,
+      "loss": 0.6158,
+      "step": 7999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7768978652343665,
+      "learning_rate": 1.764838485293966e-06,
+      "loss": 0.6817,
+      "step": 8000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9138605814810459,
+      "learning_rate": 1.7629700113284898e-06,
+      "loss": 0.6924,
+      "step": 8001
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8006453653947696,
+      "learning_rate": 1.7611024313638036e-06,
+      "loss": 0.685,
+      "step": 8002
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.621591850869959,
+      "learning_rate": 1.7592357456026065e-06,
+      "loss": 0.668,
+      "step": 8003
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5252090986604172,
+      "learning_rate": 1.7573699542475009e-06,
+      "loss": 0.6939,
+      "step": 8004
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.594839187279194,
+      "learning_rate": 1.7555050575009836e-06,
+      "loss": 0.6475,
+      "step": 8005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7306387118407995,
+      "learning_rate": 1.7536410555654658e-06,
+      "loss": 0.6955,
+      "step": 8006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7925998469644877,
+      "learning_rate": 1.7517779486432495e-06,
+      "loss": 0.691,
+      "step": 8007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8125246527297612,
+      "learning_rate": 1.7499157369365504e-06,
+      "loss": 0.6745,
+      "step": 8008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6051500991727927,
+      "learning_rate": 1.7480544206474824e-06,
+      "loss": 0.6287,
+      "step": 8009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.721327369042911,
+      "learning_rate": 1.7461939999780586e-06,
+      "loss": 0.685,
+      "step": 8010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8154002512357872,
+      "learning_rate": 1.7443344751302048e-06,
+      "loss": 0.7326,
+      "step": 8011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.734009665808617,
+      "learning_rate": 1.7424758463057356e-06,
+      "loss": 0.574,
+      "step": 8012
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5821619651938132,
+      "learning_rate": 1.7406181137063804e-06,
+      "loss": 0.5944,
+      "step": 8013
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6127780706422297,
+      "learning_rate": 1.7387612775337703e-06,
+      "loss": 0.6382,
+      "step": 8014
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7682106393097323,
+      "learning_rate": 1.7369053379894285e-06,
+      "loss": 0.7225,
+      "step": 8015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6834214431659507,
+      "learning_rate": 1.7350502952747916e-06,
+      "loss": 0.7673,
+      "step": 8016
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7819707141660162,
+      "learning_rate": 1.7331961495911997e-06,
+      "loss": 0.6721,
+      "step": 8017
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6772630357351819,
+      "learning_rate": 1.731342901139884e-06,
+      "loss": 0.6176,
+      "step": 8018
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6709942471951538,
+      "learning_rate": 1.7294905501219915e-06,
+      "loss": 0.6154,
+      "step": 8019
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7757882299108645,
+      "learning_rate": 1.7276390967385614e-06,
+      "loss": 0.758,
+      "step": 8020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7015110760484848,
+      "learning_rate": 1.7257885411905416e-06,
+      "loss": 0.7283,
+      "step": 8021
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7701017096659257,
+      "learning_rate": 1.723938883678784e-06,
+      "loss": 0.7977,
+      "step": 8022
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.93781008303018,
+      "learning_rate": 1.7220901244040355e-06,
+      "loss": 0.6425,
+      "step": 8023
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.545177983002849,
+      "learning_rate": 1.7202422635669536e-06,
+      "loss": 0.7073,
+      "step": 8024
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8987430084382586,
+      "learning_rate": 1.718395301368091e-06,
+      "loss": 0.764,
+      "step": 8025
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4753951907006009,
+      "learning_rate": 1.7165492380079084e-06,
+      "loss": 0.682,
+      "step": 8026
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8881432214443394,
+      "learning_rate": 1.7147040736867704e-06,
+      "loss": 0.7082,
+      "step": 8027
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7149537130909085,
+      "learning_rate": 1.7128598086049353e-06,
+      "loss": 0.6842,
+      "step": 8028
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.612966254277709,
+      "learning_rate": 1.711016442962573e-06,
+      "loss": 0.6647,
+      "step": 8029
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7496388997343648,
+      "learning_rate": 1.7091739769597492e-06,
+      "loss": 0.7958,
+      "step": 8030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6416089663102067,
+      "learning_rate": 1.7073324107964363e-06,
+      "loss": 0.6912,
+      "step": 8031
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7959977773068057,
+      "learning_rate": 1.7054917446725083e-06,
+      "loss": 0.7446,
+      "step": 8032
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.758838901126341,
+      "learning_rate": 1.7036519787877393e-06,
+      "loss": 0.7721,
+      "step": 8033
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6413026632094811,
+      "learning_rate": 1.701813113341806e-06,
+      "loss": 0.6721,
+      "step": 8034
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6692477027547694,
+      "learning_rate": 1.699975148534293e-06,
+      "loss": 0.5955,
+      "step": 8035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.641655359956405,
+      "learning_rate": 1.6981380845646779e-06,
+      "loss": 0.6715,
+      "step": 8036
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6286894104599299,
+      "learning_rate": 1.6963019216323472e-06,
+      "loss": 0.6614,
+      "step": 8037
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5692556277032406,
+      "learning_rate": 1.69446665993659e-06,
+      "loss": 0.7314,
+      "step": 8038
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7951697399440905,
+      "learning_rate": 1.6926322996765899e-06,
+      "loss": 0.6347,
+      "step": 8039
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.76987488315246,
+      "learning_rate": 1.6907988410514408e-06,
+      "loss": 0.7911,
+      "step": 8040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6345706578407215,
+      "learning_rate": 1.6889662842601384e-06,
+      "loss": 0.6243,
+      "step": 8041
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7310740912828402,
+      "learning_rate": 1.6871346295015744e-06,
+      "loss": 0.6846,
+      "step": 8042
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6632516322796669,
+      "learning_rate": 1.6853038769745466e-06,
+      "loss": 0.6219,
+      "step": 8043
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4446934371691147,
+      "learning_rate": 1.6834740268777594e-06,
+      "loss": 0.6495,
+      "step": 8044
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7943768358241974,
+      "learning_rate": 1.681645079409807e-06,
+      "loss": 0.7088,
+      "step": 8045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.619771756196256,
+      "learning_rate": 1.6798170347692e-06,
+      "loss": 0.7114,
+      "step": 8046
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4099576457270464,
+      "learning_rate": 1.6779898931543382e-06,
+      "loss": 0.5875,
+      "step": 8047
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7769641316748075,
+      "learning_rate": 1.6761636547635308e-06,
+      "loss": 0.6824,
+      "step": 8048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6713721303852005,
+      "learning_rate": 1.6743383197949925e-06,
+      "loss": 0.7363,
+      "step": 8049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5786940687358804,
+      "learning_rate": 1.6725138884468273e-06,
+      "loss": 0.6727,
+      "step": 8050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.78872539059089,
+      "learning_rate": 1.6706903609170522e-06,
+      "loss": 0.7712,
+      "step": 8051
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5147655598325824,
+      "learning_rate": 1.6688677374035856e-06,
+      "loss": 0.6125,
+      "step": 8052
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6544674731977558,
+      "learning_rate": 1.6670460181042381e-06,
+      "loss": 0.7031,
+      "step": 8053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.709302577018215,
+      "learning_rate": 1.6652252032167337e-06,
+      "loss": 0.6829,
+      "step": 8054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6225215150122227,
+      "learning_rate": 1.6634052929386946e-06,
+      "loss": 0.7626,
+      "step": 8055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5709590449037962,
+      "learning_rate": 1.661586287467638e-06,
+      "loss": 0.6229,
+      "step": 8056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8451733707947704,
+      "learning_rate": 1.6597681870009917e-06,
+      "loss": 0.7429,
+      "step": 8057
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.620867770846572,
+      "learning_rate": 1.6579509917360859e-06,
+      "loss": 0.6697,
+      "step": 8058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6393845757334204,
+      "learning_rate": 1.6561347018701412e-06,
+      "loss": 0.5999,
+      "step": 8059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7245146300654126,
+      "learning_rate": 1.6543193176002936e-06,
+      "loss": 0.6413,
+      "step": 8060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6501570293526682,
+      "learning_rate": 1.6525048391235698e-06,
+      "loss": 0.6118,
+      "step": 8061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5671003694062557,
+      "learning_rate": 1.6506912666369067e-06,
+      "loss": 0.7136,
+      "step": 8062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7856317394038808,
+      "learning_rate": 1.6488786003371393e-06,
+      "loss": 0.7216,
+      "step": 8063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.673438416970208,
+      "learning_rate": 1.647066840421001e-06,
+      "loss": 0.6913,
+      "step": 8064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6861994386869004,
+      "learning_rate": 1.6452559870851336e-06,
+      "loss": 0.7163,
+      "step": 8065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6480156694754329,
+      "learning_rate": 1.6434460405260733e-06,
+      "loss": 0.7316,
+      "step": 8066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.465375917084765,
+      "learning_rate": 1.6416370009402627e-06,
+      "loss": 0.6073,
+      "step": 8067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6110628327794576,
+      "learning_rate": 1.6398288685240494e-06,
+      "loss": 0.7929,
+      "step": 8068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6048075073218324,
+      "learning_rate": 1.6380216434736706e-06,
+      "loss": 0.6459,
+      "step": 8069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.643815248003964,
+      "learning_rate": 1.6362153259852775e-06,
+      "loss": 0.6194,
+      "step": 8070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6452411850619555,
+      "learning_rate": 1.6344099162549143e-06,
+      "loss": 0.7063,
+      "step": 8071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7036325914469437,
+      "learning_rate": 1.6326054144785319e-06,
+      "loss": 0.6956,
+      "step": 8072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7277501315511394,
+      "learning_rate": 1.6308018208519815e-06,
+      "loss": 0.6394,
+      "step": 8073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7138251157363322,
+      "learning_rate": 1.6289991355710121e-06,
+      "loss": 0.7295,
+      "step": 8074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6265331835084222,
+      "learning_rate": 1.6271973588312784e-06,
+      "loss": 0.5779,
+      "step": 8075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6653808390691958,
+      "learning_rate": 1.6253964908283382e-06,
+      "loss": 0.7051,
+      "step": 8076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6804402910950127,
+      "learning_rate": 1.6235965317576418e-06,
+      "loss": 0.7356,
+      "step": 8077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7370222042757812,
+      "learning_rate": 1.6217974818145532e-06,
+      "loss": 0.7099,
+      "step": 8078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4747458546716434,
+      "learning_rate": 1.6199993411943238e-06,
+      "loss": 0.5782,
+      "step": 8079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6859470291193008,
+      "learning_rate": 1.6182021100921185e-06,
+      "loss": 0.7637,
+      "step": 8080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7702922525523404,
+      "learning_rate": 1.6164057887029994e-06,
+      "loss": 0.5905,
+      "step": 8081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6709702750784812,
+      "learning_rate": 1.6146103772219246e-06,
+      "loss": 0.6451,
+      "step": 8082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6401880834646672,
+      "learning_rate": 1.612815875843763e-06,
+      "loss": 0.5818,
+      "step": 8083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8810582744200497,
+      "learning_rate": 1.611022284763274e-06,
+      "loss": 0.7115,
+      "step": 8084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7683313979746031,
+      "learning_rate": 1.6092296041751277e-06,
+      "loss": 0.7298,
+      "step": 8085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7469266290104861,
+      "learning_rate": 1.607437834273894e-06,
+      "loss": 0.7389,
+      "step": 8086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5873409497942523,
+      "learning_rate": 1.6056469752540349e-06,
+      "loss": 0.7042,
+      "step": 8087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.683437781534904,
+      "learning_rate": 1.6038570273099273e-06,
+      "loss": 0.6705,
+      "step": 8088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.847510008683372,
+      "learning_rate": 1.602067990635835e-06,
+      "loss": 0.6703,
+      "step": 8089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5865058761843016,
+      "learning_rate": 1.6002798654259333e-06,
+      "loss": 0.6765,
+      "step": 8090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4981225938372302,
+      "learning_rate": 1.5984926518742982e-06,
+      "loss": 0.5621,
+      "step": 8091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6472692118142727,
+      "learning_rate": 1.596706350174898e-06,
+      "loss": 0.6508,
+      "step": 8092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.645000625397688,
+      "learning_rate": 1.594920960521611e-06,
+      "loss": 0.756,
+      "step": 8093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.628324434163129,
+      "learning_rate": 1.5931364831082152e-06,
+      "loss": 0.7143,
+      "step": 8094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8753708051209703,
+      "learning_rate": 1.5913529181283837e-06,
+      "loss": 0.7856,
+      "step": 8095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6221303737097068,
+      "learning_rate": 1.5895702657756984e-06,
+      "loss": 0.7017,
+      "step": 8096
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.601246385274789,
+      "learning_rate": 1.5877885262436333e-06,
+      "loss": 0.6951,
+      "step": 8097
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5418810995917551,
+      "learning_rate": 1.5860076997255725e-06,
+      "loss": 0.6298,
+      "step": 8098
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7361284550203793,
+      "learning_rate": 1.5842277864147971e-06,
+      "loss": 0.7036,
+      "step": 8099
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5257967081619923,
+      "learning_rate": 1.5824487865044857e-06,
+      "loss": 0.5985,
+      "step": 8100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.83644377022114,
+      "learning_rate": 1.5806707001877253e-06,
+      "loss": 0.7191,
+      "step": 8101
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6539965105696637,
+      "learning_rate": 1.5788935276574947e-06,
+      "loss": 0.7151,
+      "step": 8102
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7310246798846824,
+      "learning_rate": 1.5771172691066793e-06,
+      "loss": 0.6899,
+      "step": 8103
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.723225438860287,
+      "learning_rate": 1.5753419247280676e-06,
+      "loss": 0.7134,
+      "step": 8104
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5959154250822916,
+      "learning_rate": 1.573567494714342e-06,
+      "loss": 0.677,
+      "step": 8105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.742594180725846,
+      "learning_rate": 1.5717939792580916e-06,
+      "loss": 0.6746,
+      "step": 8106
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.709142755507124,
+      "learning_rate": 1.5700213785518003e-06,
+      "loss": 0.7439,
+      "step": 8107
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6564530461284355,
+      "learning_rate": 1.5682496927878577e-06,
+      "loss": 0.6564,
+      "step": 8108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.949082537417003,
+      "learning_rate": 1.5664789221585552e-06,
+      "loss": 0.6938,
+      "step": 8109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7131596161923532,
+      "learning_rate": 1.5647090668560794e-06,
+      "loss": 0.6444,
+      "step": 8110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6780301599326606,
+      "learning_rate": 1.5629401270725197e-06,
+      "loss": 0.6874,
+      "step": 8111
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4744301947942198,
+      "learning_rate": 1.5611721029998716e-06,
+      "loss": 0.7268,
+      "step": 8112
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6870236838857529,
+      "learning_rate": 1.5594049948300205e-06,
+      "loss": 0.6121,
+      "step": 8113
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4872314938621514,
+      "learning_rate": 1.557638802754763e-06,
+      "loss": 0.6208,
+      "step": 8114
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7013662140599177,
+      "learning_rate": 1.5558735269657877e-06,
+      "loss": 0.6237,
+      "step": 8115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7271219737991148,
+      "learning_rate": 1.5541091676546904e-06,
+      "loss": 0.6577,
+      "step": 8116
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5893707443428777,
+      "learning_rate": 1.5523457250129648e-06,
+      "loss": 0.6867,
+      "step": 8117
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9039068155013918,
+      "learning_rate": 1.5505831992320031e-06,
+      "loss": 0.665,
+      "step": 8118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7217487973886862,
+      "learning_rate": 1.5488215905031033e-06,
+      "loss": 0.7883,
+      "step": 8119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6230993010708583,
+      "learning_rate": 1.5470608990174551e-06,
+      "loss": 0.724,
+      "step": 8120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5198966086936623,
+      "learning_rate": 1.5453011249661577e-06,
+      "loss": 0.6361,
+      "step": 8121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7769794565911328,
+      "learning_rate": 1.543542268540209e-06,
+      "loss": 0.7086,
+      "step": 8122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7323610176209234,
+      "learning_rate": 1.5417843299305002e-06,
+      "loss": 0.6249,
+      "step": 8123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6130541426278338,
+      "learning_rate": 1.5400273093278328e-06,
+      "loss": 0.5961,
+      "step": 8124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8589668993459887,
+      "learning_rate": 1.5382712069228999e-06,
+      "loss": 0.6291,
+      "step": 8125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.588078222019581,
+      "learning_rate": 1.5365160229063004e-06,
+      "loss": 0.6874,
+      "step": 8126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7482045277109108,
+      "learning_rate": 1.5347617574685358e-06,
+      "loss": 0.5672,
+      "step": 8127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6977910833089311,
+      "learning_rate": 1.5330084107999976e-06,
+      "loss": 0.6242,
+      "step": 8128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8183946031637237,
+      "learning_rate": 1.5312559830909878e-06,
+      "loss": 0.714,
+      "step": 8129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5263848763255998,
+      "learning_rate": 1.5295044745317068e-06,
+      "loss": 0.747,
+      "step": 8130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8710990805940892,
+      "learning_rate": 1.5277538853122497e-06,
+      "loss": 0.731,
+      "step": 8131
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6603387288595752,
+      "learning_rate": 1.5260042156226184e-06,
+      "loss": 0.6077,
+      "step": 8132
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.671883523275829,
+      "learning_rate": 1.5242554656527097e-06,
+      "loss": 0.8259,
+      "step": 8133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.840611579339051,
+      "learning_rate": 1.522507635592324e-06,
+      "loss": 0.609,
+      "step": 8134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.851169885764712,
+      "learning_rate": 1.5207607256311641e-06,
+      "loss": 0.7148,
+      "step": 8135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.619183101913956,
+      "learning_rate": 1.5190147359588237e-06,
+      "loss": 0.6515,
+      "step": 8136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7002117308407043,
+      "learning_rate": 1.517269666764809e-06,
+      "loss": 0.6969,
+      "step": 8137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7142362259110455,
+      "learning_rate": 1.5155255182385143e-06,
+      "loss": 0.6511,
+      "step": 8138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.601914764339988,
+      "learning_rate": 1.5137822905692423e-06,
+      "loss": 0.6475,
+      "step": 8139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7644138479405698,
+      "learning_rate": 1.5120399839461953e-06,
+      "loss": 0.6861,
+      "step": 8140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7064974699464002,
+      "learning_rate": 1.5102985985584695e-06,
+      "loss": 0.7408,
+      "step": 8141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6246618466450624,
+      "learning_rate": 1.508558134595066e-06,
+      "loss": 0.6684,
+      "step": 8142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6562685064954157,
+      "learning_rate": 1.5068185922448887e-06,
+      "loss": 0.6657,
+      "step": 8143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9585946166025932,
+      "learning_rate": 1.5050799716967313e-06,
+      "loss": 0.6331,
+      "step": 8144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.671683570356811,
+      "learning_rate": 1.5033422731392977e-06,
+      "loss": 0.7035,
+      "step": 8145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.550323061271016,
+      "learning_rate": 1.5016054967611904e-06,
+      "loss": 0.6498,
+      "step": 8146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7797696570553185,
+      "learning_rate": 1.499869642750904e-06,
+      "loss": 0.747,
+      "step": 8147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7232419058262327,
+      "learning_rate": 1.4981347112968426e-06,
+      "loss": 0.7048,
+      "step": 8148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8706299110360487,
+      "learning_rate": 1.4964007025873017e-06,
+      "loss": 0.6893,
+      "step": 8149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.604912756002985,
+      "learning_rate": 1.4946676168104834e-06,
+      "loss": 0.6486,
+      "step": 8150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7382472735836743,
+      "learning_rate": 1.4929354541544882e-06,
+      "loss": 0.7424,
+      "step": 8151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9145589612518605,
+      "learning_rate": 1.4912042148073124e-06,
+      "loss": 0.7238,
+      "step": 8152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5925140534896371,
+      "learning_rate": 1.489473898956857e-06,
+      "loss": 0.5963,
+      "step": 8153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7374229282832934,
+      "learning_rate": 1.4877445067909213e-06,
+      "loss": 0.7043,
+      "step": 8154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8373494978688998,
+      "learning_rate": 1.4860160384972e-06,
+      "loss": 0.6932,
+      "step": 8155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7972090772416303,
+      "learning_rate": 1.4842884942632952e-06,
+      "loss": 0.7046,
+      "step": 8156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9947400366471757,
+      "learning_rate": 1.4825618742767045e-06,
+      "loss": 0.7492,
+      "step": 8157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5968741451412138,
+      "learning_rate": 1.4808361787248237e-06,
+      "loss": 0.6735,
+      "step": 8158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7897791644317969,
+      "learning_rate": 1.4791114077949497e-06,
+      "loss": 0.6743,
+      "step": 8159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.525728328072972,
+      "learning_rate": 1.4773875616742828e-06,
+      "loss": 0.6981,
+      "step": 8160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8020126122351128,
+      "learning_rate": 1.4756646405499164e-06,
+      "loss": 0.7265,
+      "step": 8161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7665628087081688,
+      "learning_rate": 1.4739426446088467e-06,
+      "loss": 0.7165,
+      "step": 8162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7674743044682641,
+      "learning_rate": 1.472221574037972e-06,
+      "loss": 0.6207,
+      "step": 8163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6395601025438578,
+      "learning_rate": 1.4705014290240838e-06,
+      "loss": 0.6381,
+      "step": 8164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5560174701688523,
+      "learning_rate": 1.4687822097538796e-06,
+      "loss": 0.5278,
+      "step": 8165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5288415132268074,
+      "learning_rate": 1.4670639164139555e-06,
+      "loss": 0.6048,
+      "step": 8166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.805040181247626,
+      "learning_rate": 1.4653465491908003e-06,
+      "loss": 0.6458,
+      "step": 8167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.552499177462516,
+      "learning_rate": 1.4636301082708127e-06,
+      "loss": 0.6218,
+      "step": 8168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6076231371219833,
+      "learning_rate": 1.4619145938402813e-06,
+      "loss": 0.6282,
+      "step": 8169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7702824958542995,
+      "learning_rate": 1.4602000060853994e-06,
+      "loss": 0.7116,
+      "step": 8170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8301214132381038,
+      "learning_rate": 1.4584863451922615e-06,
+      "loss": 0.7077,
+      "step": 8171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8228686877917728,
+      "learning_rate": 1.4567736113468545e-06,
+      "loss": 0.7512,
+      "step": 8172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7104293452984662,
+      "learning_rate": 1.4550618047350728e-06,
+      "loss": 0.7394,
+      "step": 8173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8864743707993548,
+      "learning_rate": 1.4533509255427013e-06,
+      "loss": 0.7376,
+      "step": 8174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7191844775183622,
+      "learning_rate": 1.4516409739554338e-06,
+      "loss": 0.6737,
+      "step": 8175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7633488087541436,
+      "learning_rate": 1.4499319501588582e-06,
+      "loss": 0.7298,
+      "step": 8176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7316191985407565,
+      "learning_rate": 1.4482238543384596e-06,
+      "loss": 0.6715,
+      "step": 8177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5948138141341184,
+      "learning_rate": 1.4465166866796298e-06,
+      "loss": 0.6769,
+      "step": 8178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6954440427739106,
+      "learning_rate": 1.4448104473676483e-06,
+      "loss": 0.7358,
+      "step": 8179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8078562613994282,
+      "learning_rate": 1.4431051365877058e-06,
+      "loss": 0.7675,
+      "step": 8180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6272987902402685,
+      "learning_rate": 1.4414007545248875e-06,
+      "loss": 0.6328,
+      "step": 8181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6629866223470973,
+      "learning_rate": 1.4396973013641736e-06,
+      "loss": 0.7112,
+      "step": 8182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7560931270102156,
+      "learning_rate": 1.4379947772904502e-06,
+      "loss": 0.6473,
+      "step": 8183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6533183558470694,
+      "learning_rate": 1.4362931824884995e-06,
+      "loss": 0.7396,
+      "step": 8184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7162203646747494,
+      "learning_rate": 1.4345925171430019e-06,
+      "loss": 0.6405,
+      "step": 8185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7371762319524415,
+      "learning_rate": 1.4328927814385397e-06,
+      "loss": 0.6946,
+      "step": 8186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6011820265500123,
+      "learning_rate": 1.4311939755595904e-06,
+      "loss": 0.6324,
+      "step": 8187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.718557478458414,
+      "learning_rate": 1.4294960996905328e-06,
+      "loss": 0.7573,
+      "step": 8188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8507287971729904,
+      "learning_rate": 1.4277991540156487e-06,
+      "loss": 0.6725,
+      "step": 8189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6812252065295867,
+      "learning_rate": 1.4261031387191093e-06,
+      "loss": 0.6939,
+      "step": 8190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5990936705216816,
+      "learning_rate": 1.424408053984997e-06,
+      "loss": 0.6721,
+      "step": 8191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.644941347595725,
+      "learning_rate": 1.4227138999972801e-06,
+      "loss": 0.6082,
+      "step": 8192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6488723534423682,
+      "learning_rate": 1.421020676939835e-06,
+      "loss": 0.6637,
+      "step": 8193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.641270793531587,
+      "learning_rate": 1.4193283849964379e-06,
+      "loss": 0.7056,
+      "step": 8194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6060072990891627,
+      "learning_rate": 1.4176370243507564e-06,
+      "loss": 0.6986,
+      "step": 8195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7266934811155912,
+      "learning_rate": 1.4159465951863638e-06,
+      "loss": 0.7088,
+      "step": 8196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5718313566278994,
+      "learning_rate": 1.4142570976867276e-06,
+      "loss": 0.7139,
+      "step": 8197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7411643009476758,
+      "learning_rate": 1.4125685320352167e-06,
+      "loss": 0.6195,
+      "step": 8198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8999500432719307,
+      "learning_rate": 1.4108808984151023e-06,
+      "loss": 0.7139,
+      "step": 8199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.493831320737935,
+      "learning_rate": 1.4091941970095446e-06,
+      "loss": 0.6281,
+      "step": 8200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7084876257684942,
+      "learning_rate": 1.4075084280016138e-06,
+      "loss": 0.7867,
+      "step": 8201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7936129382568762,
+      "learning_rate": 1.4058235915742702e-06,
+      "loss": 0.7075,
+      "step": 8202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6964101364789455,
+      "learning_rate": 1.4041396879103763e-06,
+      "loss": 0.6834,
+      "step": 8203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.738053483920836,
+      "learning_rate": 1.4024567171926984e-06,
+      "loss": 0.7705,
+      "step": 8204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8014855759132964,
+      "learning_rate": 1.4007746796038912e-06,
+      "loss": 0.6916,
+      "step": 8205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7326603410539312,
+      "learning_rate": 1.3990935753265155e-06,
+      "loss": 0.816,
+      "step": 8206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7264773657111878,
+      "learning_rate": 1.397413404543031e-06,
+      "loss": 0.7231,
+      "step": 8207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5624102154971813,
+      "learning_rate": 1.3957341674357904e-06,
+      "loss": 0.7574,
+      "step": 8208
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5545206389114075,
+      "learning_rate": 1.3940558641870517e-06,
+      "loss": 0.6499,
+      "step": 8209
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6880268153759315,
+      "learning_rate": 1.3923784949789654e-06,
+      "loss": 0.7573,
+      "step": 8210
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.720853432426547,
+      "learning_rate": 1.390702059993585e-06,
+      "loss": 0.7196,
+      "step": 8211
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6481869374763833,
+      "learning_rate": 1.3890265594128638e-06,
+      "loss": 0.6756,
+      "step": 8212
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6094648483301166,
+      "learning_rate": 1.3873519934186474e-06,
+      "loss": 0.7078,
+      "step": 8213
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.618792032977977,
+      "learning_rate": 1.3856783621926873e-06,
+      "loss": 0.682,
+      "step": 8214
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5878214192541364,
+      "learning_rate": 1.3840056659166257e-06,
+      "loss": 0.7473,
+      "step": 8215
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5842895543466722,
+      "learning_rate": 1.3823339047720096e-06,
+      "loss": 0.6335,
+      "step": 8216
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7782310880484486,
+      "learning_rate": 1.3806630789402852e-06,
+      "loss": 0.7653,
+      "step": 8217
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6956498406697338,
+      "learning_rate": 1.3789931886027907e-06,
+      "loss": 0.5894,
+      "step": 8218
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7475029519745575,
+      "learning_rate": 1.37732423394077e-06,
+      "loss": 0.7174,
+      "step": 8219
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6154412892748924,
+      "learning_rate": 1.375656215135358e-06,
+      "loss": 0.6459,
+      "step": 8220
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9434901904421764,
+      "learning_rate": 1.3739891323675935e-06,
+      "loss": 0.6698,
+      "step": 8221
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6567625078813784,
+      "learning_rate": 1.3723229858184162e-06,
+      "loss": 0.6578,
+      "step": 8222
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.776408853955685,
+      "learning_rate": 1.3706577756686545e-06,
+      "loss": 0.7233,
+      "step": 8223
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8514879954569197,
+      "learning_rate": 1.368993502099043e-06,
+      "loss": 0.7251,
+      "step": 8224
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5768850107459567,
+      "learning_rate": 1.367330165290215e-06,
+      "loss": 0.6744,
+      "step": 8225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4427110331792115,
+      "learning_rate": 1.3656677654226957e-06,
+      "loss": 0.5715,
+      "step": 8226
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7838598756278934,
+      "learning_rate": 1.364006302676918e-06,
+      "loss": 0.7783,
+      "step": 8227
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6732822161714365,
+      "learning_rate": 1.3623457772332005e-06,
+      "loss": 0.6944,
+      "step": 8228
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7225981120830656,
+      "learning_rate": 1.3606861892717715e-06,
+      "loss": 0.5786,
+      "step": 8229
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7587114484153354,
+      "learning_rate": 1.359027538972756e-06,
+      "loss": 0.7033,
+      "step": 8230
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.850649807807029,
+      "learning_rate": 1.3573698265161683e-06,
+      "loss": 0.6353,
+      "step": 8231
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6641610341874145,
+      "learning_rate": 1.3557130520819328e-06,
+      "loss": 0.7124,
+      "step": 8232
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9149203507188295,
+      "learning_rate": 1.3540572158498622e-06,
+      "loss": 0.7167,
+      "step": 8233
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7525317054907879,
+      "learning_rate": 1.3524023179996725e-06,
+      "loss": 0.767,
+      "step": 8234
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7889664661142302,
+      "learning_rate": 1.3507483587109805e-06,
+      "loss": 0.7336,
+      "step": 8235
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5378168604185913,
+      "learning_rate": 1.3490953381632933e-06,
+      "loss": 0.6962,
+      "step": 8236
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8147784384972894,
+      "learning_rate": 1.347443256536024e-06,
+      "loss": 0.6153,
+      "step": 8237
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7302525034904064,
+      "learning_rate": 1.3457921140084761e-06,
+      "loss": 0.7129,
+      "step": 8238
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5835440184825371,
+      "learning_rate": 1.3441419107598575e-06,
+      "loss": 0.7461,
+      "step": 8239
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9160856857123725,
+      "learning_rate": 1.3424926469692734e-06,
+      "loss": 0.7012,
+      "step": 8240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7432989014314013,
+      "learning_rate": 1.340844322815723e-06,
+      "loss": 0.698,
+      "step": 8241
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.826200832807161,
+      "learning_rate": 1.3391969384781066e-06,
+      "loss": 0.768,
+      "step": 8242
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6947304695110976,
+      "learning_rate": 1.3375504941352257e-06,
+      "loss": 0.666,
+      "step": 8243
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6355678464956755,
+      "learning_rate": 1.3359049899657706e-06,
+      "loss": 0.6915,
+      "step": 8244
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7745025367531957,
+      "learning_rate": 1.3342604261483406e-06,
+      "loss": 0.7305,
+      "step": 8245
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7418465569279045,
+      "learning_rate": 1.3326168028614206e-06,
+      "loss": 0.6738,
+      "step": 8246
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7872437432117918,
+      "learning_rate": 1.3309741202834047e-06,
+      "loss": 0.6452,
+      "step": 8247
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7614011552562099,
+      "learning_rate": 1.3293323785925816e-06,
+      "loss": 0.6973,
+      "step": 8248
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8171301925702645,
+      "learning_rate": 1.327691577967133e-06,
+      "loss": 0.6592,
+      "step": 8249
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7057564407848962,
+      "learning_rate": 1.326051718585144e-06,
+      "loss": 0.7125,
+      "step": 8250
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9809910038871592,
+      "learning_rate": 1.324412800624597e-06,
+      "loss": 0.6787,
+      "step": 8251
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6486996929916975,
+      "learning_rate": 1.3227748242633675e-06,
+      "loss": 0.6453,
+      "step": 8252
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.612309839141793,
+      "learning_rate": 1.3211377896792365e-06,
+      "loss": 0.6477,
+      "step": 8253
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7284065002363935,
+      "learning_rate": 1.319501697049874e-06,
+      "loss": 0.6668,
+      "step": 8254
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7727740544108945,
+      "learning_rate": 1.3178665465528551e-06,
+      "loss": 0.6749,
+      "step": 8255
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7506034964250732,
+      "learning_rate": 1.3162323383656506e-06,
+      "loss": 0.5859,
+      "step": 8256
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4653029512735753,
+      "learning_rate": 1.3145990726656244e-06,
+      "loss": 0.6423,
+      "step": 8257
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7090647923838687,
+      "learning_rate": 1.3129667496300446e-06,
+      "loss": 0.6786,
+      "step": 8258
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7074457410507926,
+      "learning_rate": 1.3113353694360764e-06,
+      "loss": 0.6967,
+      "step": 8259
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.469735361171108,
+      "learning_rate": 1.309704932260777e-06,
+      "loss": 0.6104,
+      "step": 8260
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7563334951394989,
+      "learning_rate": 1.3080754382811055e-06,
+      "loss": 0.7081,
+      "step": 8261
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9075491493123156,
+      "learning_rate": 1.30644688767392e-06,
+      "loss": 0.769,
+      "step": 8262
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.911303763578774,
+      "learning_rate": 1.3048192806159721e-06,
+      "loss": 0.598,
+      "step": 8263
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7526459129348975,
+      "learning_rate": 1.3031926172839126e-06,
+      "loss": 0.7159,
+      "step": 8264
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6665938973477412,
+      "learning_rate": 1.301566897854295e-06,
+      "loss": 0.6275,
+      "step": 8265
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5753078101727518,
+      "learning_rate": 1.2999421225035602e-06,
+      "loss": 0.634,
+      "step": 8266
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.537542156790524,
+      "learning_rate": 1.298318291408054e-06,
+      "loss": 0.602,
+      "step": 8267
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8364242230592978,
+      "learning_rate": 1.2966954047440194e-06,
+      "loss": 0.6858,
+      "step": 8268
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5620594579846787,
+      "learning_rate": 1.2950734626875927e-06,
+      "loss": 0.6455,
+      "step": 8269
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.643541532600793,
+      "learning_rate": 1.2934524654148118e-06,
+      "loss": 0.6662,
+      "step": 8270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7780861869537536,
+      "learning_rate": 1.2918324131016135e-06,
+      "loss": 0.6691,
+      "step": 8271
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.477593720782914,
+      "learning_rate": 1.290213305923823e-06,
+      "loss": 0.6359,
+      "step": 8272
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.873980085787594,
+      "learning_rate": 1.2885951440571754e-06,
+      "loss": 0.6693,
+      "step": 8273
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6247128817864538,
+      "learning_rate": 1.286977927677291e-06,
+      "loss": 0.6256,
+      "step": 8274
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.723425124139416,
+      "learning_rate": 1.285361656959696e-06,
+      "loss": 0.7455,
+      "step": 8275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5730964779506738,
+      "learning_rate": 1.2837463320798138e-06,
+      "loss": 0.7081,
+      "step": 8276
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.566190339493104,
+      "learning_rate": 1.2821319532129584e-06,
+      "loss": 0.6472,
+      "step": 8277
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.870457612150168,
+      "learning_rate": 1.2805185205343462e-06,
+      "loss": 0.6416,
+      "step": 8278
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.63866119536756,
+      "learning_rate": 1.278906034219094e-06,
+      "loss": 0.7834,
+      "step": 8279
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8807338285351183,
+      "learning_rate": 1.2772944944422073e-06,
+      "loss": 0.721,
+      "step": 8280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8934271455408902,
+      "learning_rate": 1.2756839013785971e-06,
+      "loss": 0.7553,
+      "step": 8281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9750340584837798,
+      "learning_rate": 1.274074255203065e-06,
+      "loss": 0.7215,
+      "step": 8282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.605310988428569,
+      "learning_rate": 1.2724655560903133e-06,
+      "loss": 0.7834,
+      "step": 8283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.695518140987458,
+      "learning_rate": 1.2708578042149444e-06,
+      "loss": 0.7367,
+      "step": 8284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.668408369362041,
+      "learning_rate": 1.2692509997514513e-06,
+      "loss": 0.5848,
+      "step": 8285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7085874099620402,
+      "learning_rate": 1.2676451428742297e-06,
+      "loss": 0.744,
+      "step": 8286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8688038523063668,
+      "learning_rate": 1.2660402337575672e-06,
+      "loss": 0.681,
+      "step": 8287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.969306422104363,
+      "learning_rate": 1.2644362725756531e-06,
+      "loss": 0.7181,
+      "step": 8288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6681407177929888,
+      "learning_rate": 1.262833259502575e-06,
+      "loss": 0.7593,
+      "step": 8289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.661293988860556,
+      "learning_rate": 1.2612311947123102e-06,
+      "loss": 0.7058,
+      "step": 8290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.840624486989448,
+      "learning_rate": 1.259630078378743e-06,
+      "loss": 0.6597,
+      "step": 8291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7114529523092323,
+      "learning_rate": 1.2580299106756444e-06,
+      "loss": 0.6687,
+      "step": 8292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.579609069103059,
+      "learning_rate": 1.2564306917766888e-06,
+      "loss": 0.6612,
+      "step": 8293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6687574529398108,
+      "learning_rate": 1.2548324218554508e-06,
+      "loss": 0.6822,
+      "step": 8294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6662587476244746,
+      "learning_rate": 1.2532351010853916e-06,
+      "loss": 0.701,
+      "step": 8295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.780952607420748,
+      "learning_rate": 1.251638729639878e-06,
+      "loss": 0.6609,
+      "step": 8296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6701796692628006,
+      "learning_rate": 1.2500433076921737e-06,
+      "loss": 0.7236,
+      "step": 8297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6958271656639399,
+      "learning_rate": 1.2484488354154322e-06,
+      "loss": 0.7342,
+      "step": 8298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5776014698998455,
+      "learning_rate": 1.2468553129827132e-06,
+      "loss": 0.65,
+      "step": 8299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7931854484696272,
+      "learning_rate": 1.2452627405669637e-06,
+      "loss": 0.8245,
+      "step": 8300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7502794056969053,
+      "learning_rate": 1.2436711183410344e-06,
+      "loss": 0.5966,
+      "step": 8301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6829815801088328,
+      "learning_rate": 1.2420804464776736e-06,
+      "loss": 0.6106,
+      "step": 8302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.519296170317051,
+      "learning_rate": 1.24049072514952e-06,
+      "loss": 0.6843,
+      "step": 8303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6324457551020213,
+      "learning_rate": 1.238901954529117e-06,
+      "loss": 0.6252,
+      "step": 8304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5249173181106488,
+      "learning_rate": 1.2373141347888973e-06,
+      "loss": 0.6963,
+      "step": 8305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.622005851140321,
+      "learning_rate": 1.2357272661011943e-06,
+      "loss": 0.7062,
+      "step": 8306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8225506321546634,
+      "learning_rate": 1.2341413486382404e-06,
+      "loss": 0.6429,
+      "step": 8307
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4583705955133515,
+      "learning_rate": 1.2325563825721587e-06,
+      "loss": 0.6684,
+      "step": 8308
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7633612538154486,
+      "learning_rate": 1.2309723680749763e-06,
+      "loss": 0.6642,
+      "step": 8309
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8176392806658217,
+      "learning_rate": 1.229389305318609e-06,
+      "loss": 0.7658,
+      "step": 8310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7129799594557809,
+      "learning_rate": 1.2278071944748748e-06,
+      "loss": 0.7848,
+      "step": 8311
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6597211231973432,
+      "learning_rate": 1.226226035715491e-06,
+      "loss": 0.7077,
+      "step": 8312
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.509184909464347,
+      "learning_rate": 1.2246458292120621e-06,
+      "loss": 0.5519,
+      "step": 8313
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7605713217212096,
+      "learning_rate": 1.2230665751360983e-06,
+      "loss": 0.691,
+      "step": 8314
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6014344717266955,
+      "learning_rate": 1.2214882736590029e-06,
+      "loss": 0.6665,
+      "step": 8315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7419877348732173,
+      "learning_rate": 1.2199109249520724e-06,
+      "loss": 0.6899,
+      "step": 8316
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.726432291563157,
+      "learning_rate": 1.2183345291865089e-06,
+      "loss": 0.6821,
+      "step": 8317
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8575890894392022,
+      "learning_rate": 1.2167590865333988e-06,
+      "loss": 0.7333,
+      "step": 8318
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5533278448200125,
+      "learning_rate": 1.2151845971637365e-06,
+      "loss": 0.6554,
+      "step": 8319
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7866467208283314,
+      "learning_rate": 1.2136110612484086e-06,
+      "loss": 0.7351,
+      "step": 8320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8386786003634514,
+      "learning_rate": 1.2120384789581953e-06,
+      "loss": 0.7588,
+      "step": 8321
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5810040818908622,
+      "learning_rate": 1.2104668504637773e-06,
+      "loss": 0.6545,
+      "step": 8322
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.562322712822364,
+      "learning_rate": 1.2088961759357287e-06,
+      "loss": 0.6201,
+      "step": 8323
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4897774268415849,
+      "learning_rate": 1.207326455544522e-06,
+      "loss": 0.6774,
+      "step": 8324
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.723186570192101,
+      "learning_rate": 1.2057576894605294e-06,
+      "loss": 0.7934,
+      "step": 8325
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6427088516879875,
+      "learning_rate": 1.2041898778540096e-06,
+      "loss": 0.6514,
+      "step": 8326
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6999594989840934,
+      "learning_rate": 1.2026230208951307e-06,
+      "loss": 0.7262,
+      "step": 8327
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5443055445751726,
+      "learning_rate": 1.2010571187539454e-06,
+      "loss": 0.8429,
+      "step": 8328
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7848686000066742,
+      "learning_rate": 1.1994921716004093e-06,
+      "loss": 0.6515,
+      "step": 8329
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7576592002961013,
+      "learning_rate": 1.1979281796043752e-06,
+      "loss": 0.7112,
+      "step": 8330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7057954207529185,
+      "learning_rate": 1.1963651429355871e-06,
+      "loss": 0.6285,
+      "step": 8331
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8488761820970288,
+      "learning_rate": 1.1948030617636885e-06,
+      "loss": 0.6229,
+      "step": 8332
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8905900773570874,
+      "learning_rate": 1.1932419362582215e-06,
+      "loss": 0.8246,
+      "step": 8333
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8718436174150623,
+      "learning_rate": 1.1916817665886183e-06,
+      "loss": 0.6485,
+      "step": 8334
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.912184871813602,
+      "learning_rate": 1.1901225529242145e-06,
+      "loss": 0.7346,
+      "step": 8335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7569317956831032,
+      "learning_rate": 1.1885642954342347e-06,
+      "loss": 0.8115,
+      "step": 8336
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5771053460843896,
+      "learning_rate": 1.187006994287806e-06,
+      "loss": 0.6744,
+      "step": 8337
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8762386730077556,
+      "learning_rate": 1.1854506496539485e-06,
+      "loss": 0.6746,
+      "step": 8338
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.664766180889851,
+      "learning_rate": 1.1838952617015786e-06,
+      "loss": 0.7055,
+      "step": 8339
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8189649427529284,
+      "learning_rate": 1.182340830599511e-06,
+      "loss": 0.7021,
+      "step": 8340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5837208446038582,
+      "learning_rate": 1.1807873565164507e-06,
+      "loss": 0.5835,
+      "step": 8341
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.592398960952057,
+      "learning_rate": 1.1792348396210064e-06,
+      "loss": 0.6298,
+      "step": 8342
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7031410402159544,
+      "learning_rate": 1.1776832800816807e-06,
+      "loss": 0.6705,
+      "step": 8343
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.39023837442245,
+      "learning_rate": 1.176132678066868e-06,
+      "loss": 0.5857,
+      "step": 8344
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5811862674391624,
+      "learning_rate": 1.174583033744865e-06,
+      "loss": 0.6163,
+      "step": 8345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7060951684814682,
+      "learning_rate": 1.1730343472838568e-06,
+      "loss": 0.7131,
+      "step": 8346
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.672908454020415,
+      "learning_rate": 1.1714866188519325e-06,
+      "loss": 0.6297,
+      "step": 8347
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6320128241024243,
+      "learning_rate": 1.1699398486170755e-06,
+      "loss": 0.5627,
+      "step": 8348
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6833927256365844,
+      "learning_rate": 1.1683940367471592e-06,
+      "loss": 0.6213,
+      "step": 8349
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.722046538171397,
+      "learning_rate": 1.1668491834099606e-06,
+      "loss": 0.7279,
+      "step": 8350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6351716824061726,
+      "learning_rate": 1.1653052887731465e-06,
+      "loss": 0.5468,
+      "step": 8351
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9516902571195383,
+      "learning_rate": 1.1637623530042842e-06,
+      "loss": 0.6836,
+      "step": 8352
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5973794419071494,
+      "learning_rate": 1.1622203762708374e-06,
+      "loss": 0.6042,
+      "step": 8353
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7871573535980738,
+      "learning_rate": 1.1606793587401588e-06,
+      "loss": 0.7547,
+      "step": 8354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.593337729041243,
+      "learning_rate": 1.1591393005795049e-06,
+      "loss": 0.6016,
+      "step": 8355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6813138223461972,
+      "learning_rate": 1.1576002019560261e-06,
+      "loss": 0.6707,
+      "step": 8356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.782124641021241,
+      "learning_rate": 1.1560620630367635e-06,
+      "loss": 0.6561,
+      "step": 8357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6552514721499414,
+      "learning_rate": 1.1545248839886624e-06,
+      "loss": 0.6258,
+      "step": 8358
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7510283040904677,
+      "learning_rate": 1.152988664978556e-06,
+      "loss": 0.6952,
+      "step": 8359
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6683898421711663,
+      "learning_rate": 1.1514534061731774e-06,
+      "loss": 0.6231,
+      "step": 8360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7258501235344073,
+      "learning_rate": 1.149919107739158e-06,
+      "loss": 0.6234,
+      "step": 8361
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5408850111580221,
+      "learning_rate": 1.1483857698430178e-06,
+      "loss": 0.6398,
+      "step": 8362
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6650248318563607,
+      "learning_rate": 1.1468533926511783e-06,
+      "loss": 0.7315,
+      "step": 8363
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6331762688693068,
+      "learning_rate": 1.1453219763299572e-06,
+      "loss": 0.7218,
+      "step": 8364
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6692634858355666,
+      "learning_rate": 1.143791521045562e-06,
+      "loss": 0.662,
+      "step": 8365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7281442159501406,
+      "learning_rate": 1.1422620269641015e-06,
+      "loss": 0.6767,
+      "step": 8366
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8944519361838665,
+      "learning_rate": 1.1407334942515801e-06,
+      "loss": 0.7541,
+      "step": 8367
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.72952315228512,
+      "learning_rate": 1.139205923073894e-06,
+      "loss": 0.7566,
+      "step": 8368
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.615345607547421,
+      "learning_rate": 1.1376793135968357e-06,
+      "loss": 0.694,
+      "step": 8369
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5608929827737215,
+      "learning_rate": 1.1361536659861005e-06,
+      "loss": 0.7242,
+      "step": 8370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.814474961764568,
+      "learning_rate": 1.1346289804072664e-06,
+      "loss": 0.7095,
+      "step": 8371
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6189912166838378,
+      "learning_rate": 1.1331052570258183e-06,
+      "loss": 0.7125,
+      "step": 8372
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7259453310980772,
+      "learning_rate": 1.131582496007133e-06,
+      "loss": 0.7864,
+      "step": 8373
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5270498544236337,
+      "learning_rate": 1.1300606975164807e-06,
+      "loss": 0.6883,
+      "step": 8374
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9169050697498091,
+      "learning_rate": 1.128539861719028e-06,
+      "loss": 0.8026,
+      "step": 8375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7362936142752212,
+      "learning_rate": 1.1270199887798417e-06,
+      "loss": 0.6875,
+      "step": 8376
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6440481486462275,
+      "learning_rate": 1.1255010788638753e-06,
+      "loss": 0.6966,
+      "step": 8377
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.646675454268616,
+      "learning_rate": 1.1239831321359862e-06,
+      "loss": 0.6603,
+      "step": 8378
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.725336742038386,
+      "learning_rate": 1.1224661487609234e-06,
+      "loss": 0.7253,
+      "step": 8379
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.995017255507201,
+      "learning_rate": 1.1209501289033298e-06,
+      "loss": 0.732,
+      "step": 8380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7151332610577539,
+      "learning_rate": 1.1194350727277493e-06,
+      "loss": 0.6279,
+      "step": 8381
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.488836247487096,
+      "learning_rate": 1.1179209803986124e-06,
+      "loss": 0.5818,
+      "step": 8382
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7011228341231284,
+      "learning_rate": 1.1164078520802535e-06,
+      "loss": 0.7556,
+      "step": 8383
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5593983291258506,
+      "learning_rate": 1.1148956879369e-06,
+      "loss": 0.6098,
+      "step": 8384
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5765603798333052,
+      "learning_rate": 1.1133844881326706e-06,
+      "loss": 0.6282,
+      "step": 8385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.480114990831892,
+      "learning_rate": 1.111874252831585e-06,
+      "loss": 0.4523,
+      "step": 8386
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6096880954934176,
+      "learning_rate": 1.1103649821975527e-06,
+      "loss": 0.6659,
+      "step": 8387
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7778688660947681,
+      "learning_rate": 1.1088566763943843e-06,
+      "loss": 0.7247,
+      "step": 8388
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.803483749854071,
+      "learning_rate": 1.1073493355857823e-06,
+      "loss": 0.6551,
+      "step": 8389
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6709014820841264,
+      "learning_rate": 1.1058429599353426e-06,
+      "loss": 0.6616,
+      "step": 8390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9903915640329826,
+      "learning_rate": 1.104337549606561e-06,
+      "loss": 0.8037,
+      "step": 8391
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5789512190681958,
+      "learning_rate": 1.1028331047628282e-06,
+      "loss": 0.7054,
+      "step": 8392
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7115052153277104,
+      "learning_rate": 1.1013296255674233e-06,
+      "loss": 0.7137,
+      "step": 8393
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5073715688022733,
+      "learning_rate": 1.0998271121835302e-06,
+      "loss": 0.595,
+      "step": 8394
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.887979213898425,
+      "learning_rate": 1.0983255647742185e-06,
+      "loss": 0.707,
+      "step": 8395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.701860423331792,
+      "learning_rate": 1.0968249835024603e-06,
+      "loss": 0.7961,
+      "step": 8396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.619199574800476,
+      "learning_rate": 1.0953253685311227e-06,
+      "loss": 0.5768,
+      "step": 8397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.987229741846533,
+      "learning_rate": 1.0938267200229603e-06,
+      "loss": 0.6983,
+      "step": 8398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8108338151674153,
+      "learning_rate": 1.0923290381406316e-06,
+      "loss": 0.7066,
+      "step": 8399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7244286465155514,
+      "learning_rate": 1.0908323230466845e-06,
+      "loss": 0.6901,
+      "step": 8400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6688581986336752,
+      "learning_rate": 1.0893365749035646e-06,
+      "loss": 0.6577,
+      "step": 8401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.601690834802701,
+      "learning_rate": 1.0878417938736142e-06,
+      "loss": 0.6563,
+      "step": 8402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.653747397467601,
+      "learning_rate": 1.0863479801190645e-06,
+      "loss": 0.7102,
+      "step": 8403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6179264464657146,
+      "learning_rate": 1.0848551338020497e-06,
+      "loss": 0.702,
+      "step": 8404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6147101757367943,
+      "learning_rate": 1.0833632550845907e-06,
+      "loss": 0.6662,
+      "step": 8405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8821238740303694,
+      "learning_rate": 1.0818723441286105e-06,
+      "loss": 0.7495,
+      "step": 8406
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6447027933493943,
+      "learning_rate": 1.0803824010959253e-06,
+      "loss": 0.7016,
+      "step": 8407
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7410971803330708,
+      "learning_rate": 1.078893426148241e-06,
+      "loss": 0.6797,
+      "step": 8408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.609515505448147,
+      "learning_rate": 1.0774054194471638e-06,
+      "loss": 0.6954,
+      "step": 8409
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8647808273584001,
+      "learning_rate": 1.0759183811541973e-06,
+      "loss": 0.7015,
+      "step": 8410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.656207416925414,
+      "learning_rate": 1.074432311430732e-06,
+      "loss": 0.7844,
+      "step": 8411
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5818580586379434,
+      "learning_rate": 1.0729472104380601e-06,
+      "loss": 0.6853,
+      "step": 8412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6628763789830707,
+      "learning_rate": 1.0714630783373636e-06,
+      "loss": 0.7416,
+      "step": 8413
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8792780484068097,
+      "learning_rate": 1.0699799152897238e-06,
+      "loss": 0.8003,
+      "step": 8414
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6385320775186256,
+      "learning_rate": 1.0684977214561155e-06,
+      "loss": 0.6372,
+      "step": 8415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3279034371879015,
+      "learning_rate": 1.067016496997404e-06,
+      "loss": 0.5391,
+      "step": 8416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7867343700510006,
+      "learning_rate": 1.0655362420743587e-06,
+      "loss": 0.7462,
+      "step": 8417
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5461375040632466,
+      "learning_rate": 1.0640569568476323e-06,
+      "loss": 0.5684,
+      "step": 8418
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6583070609877475,
+      "learning_rate": 1.0625786414777805e-06,
+      "loss": 0.6747,
+      "step": 8419
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.703374935104682,
+      "learning_rate": 1.0611012961252543e-06,
+      "loss": 0.755,
+      "step": 8420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.66739442848443,
+      "learning_rate": 1.0596249209503906e-06,
+      "loss": 0.8473,
+      "step": 8421
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8210961467259086,
+      "learning_rate": 1.0581495161134325e-06,
+      "loss": 0.6847,
+      "step": 8422
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5643904053537698,
+      "learning_rate": 1.0566750817745076e-06,
+      "loss": 0.6154,
+      "step": 8423
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6738856312856583,
+      "learning_rate": 1.0552016180936442e-06,
+      "loss": 0.648,
+      "step": 8424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7984497698344761,
+      "learning_rate": 1.0537291252307668e-06,
+      "loss": 0.7255,
+      "step": 8425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.611425019120535,
+      "learning_rate": 1.0522576033456865e-06,
+      "loss": 0.5955,
+      "step": 8426
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5985930161310276,
+      "learning_rate": 1.0507870525981166e-06,
+      "loss": 0.647,
+      "step": 8427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.602373516978655,
+      "learning_rate": 1.0493174731476641e-06,
+      "loss": 0.6878,
+      "step": 8428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.684892750947452,
+      "learning_rate": 1.0478488651538244e-06,
+      "loss": 0.6671,
+      "step": 8429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.751413082663372,
+      "learning_rate": 1.0463812287759967e-06,
+      "loss": 0.8063,
+      "step": 8430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6046268645591264,
+      "learning_rate": 1.0449145641734648e-06,
+      "loss": 0.5918,
+      "step": 8431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7159864744314377,
+      "learning_rate": 1.0434488715054158e-06,
+      "loss": 0.6877,
+      "step": 8432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.631233184520412,
+      "learning_rate": 1.0419841509309293e-06,
+      "loss": 0.7765,
+      "step": 8433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.532792792300549,
+      "learning_rate": 1.0405204026089732e-06,
+      "loss": 0.5547,
+      "step": 8434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6352150853995833,
+      "learning_rate": 1.0390576266984186e-06,
+      "loss": 0.785,
+      "step": 8435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.621107971810868,
+      "learning_rate": 1.0375958233580241e-06,
+      "loss": 0.6549,
+      "step": 8436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7423638741067684,
+      "learning_rate": 1.0361349927464459e-06,
+      "loss": 0.671,
+      "step": 8437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6553449685180674,
+      "learning_rate": 1.0346751350222373e-06,
+      "loss": 0.7222,
+      "step": 8438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.69168484776965,
+      "learning_rate": 1.0332162503438382e-06,
+      "loss": 0.7429,
+      "step": 8439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7137645740993,
+      "learning_rate": 1.031758338869593e-06,
+      "loss": 0.7141,
+      "step": 8440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7547540988136974,
+      "learning_rate": 1.0303014007577306e-06,
+      "loss": 0.6966,
+      "step": 8441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6525465199540916,
+      "learning_rate": 1.0288454361663802e-06,
+      "loss": 0.6479,
+      "step": 8442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5072249642946518,
+      "learning_rate": 1.0273904452535666e-06,
+      "loss": 0.6614,
+      "step": 8443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5417080513159942,
+      "learning_rate": 1.0259364281772023e-06,
+      "loss": 0.6281,
+      "step": 8444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.586239143977994,
+      "learning_rate": 1.024483385095101e-06,
+      "loss": 0.739,
+      "step": 8445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6299264084653204,
+      "learning_rate": 1.0230313161649674e-06,
+      "loss": 0.71,
+      "step": 8446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.652423537669246,
+      "learning_rate": 1.0215802215443993e-06,
+      "loss": 0.6331,
+      "step": 8447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.486458058211073,
+      "learning_rate": 1.0201301013908926e-06,
+      "loss": 0.6625,
+      "step": 8448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7202949780514047,
+      "learning_rate": 1.0186809558618327e-06,
+      "loss": 0.7799,
+      "step": 8449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.851452722138668,
+      "learning_rate": 1.0172327851145026e-06,
+      "loss": 0.713,
+      "step": 8450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7589655012461458,
+      "learning_rate": 1.0157855893060809e-06,
+      "loss": 0.706,
+      "step": 8451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7341879348494615,
+      "learning_rate": 1.0143393685936342e-06,
+      "loss": 0.7593,
+      "step": 8452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9346470924362966,
+      "learning_rate": 1.0128941231341305e-06,
+      "loss": 0.7522,
+      "step": 8453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6773163911161024,
+      "learning_rate": 1.0114498530844253e-06,
+      "loss": 0.732,
+      "step": 8454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6882931865366573,
+      "learning_rate": 1.010006558601274e-06,
+      "loss": 0.6606,
+      "step": 8455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6757261676598967,
+      "learning_rate": 1.0085642398413243e-06,
+      "loss": 0.5838,
+      "step": 8456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8452775158196029,
+      "learning_rate": 1.0071228969611146e-06,
+      "loss": 0.7233,
+      "step": 8457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5740820266658542,
+      "learning_rate": 1.0056825301170836e-06,
+      "loss": 0.5603,
+      "step": 8458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6426284893961631,
+      "learning_rate": 1.0042431394655562e-06,
+      "loss": 0.671,
+      "step": 8459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6699412335146264,
+      "learning_rate": 1.0028047251627583e-06,
+      "loss": 0.6582,
+      "step": 8460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6824431706589318,
+      "learning_rate": 1.0013672873648083e-06,
+      "loss": 0.7196,
+      "step": 8461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.963283484565238,
+      "learning_rate": 9.999308262277152e-07,
+      "loss": 0.7811,
+      "step": 8462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5592969312288327,
+      "learning_rate": 9.98495341907385e-07,
+      "loss": 0.5841,
+      "step": 8463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.578809012565993,
+      "learning_rate": 9.9706083455962e-07,
+      "loss": 0.6566,
+      "step": 8464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7662396613126716,
+      "learning_rate": 9.9562730434011e-07,
+      "loss": 0.7146,
+      "step": 8465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.564066354381204,
+      "learning_rate": 9.941947514044437e-07,
+      "loss": 0.6503,
+      "step": 8466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.74140264632033,
+      "learning_rate": 9.927631759081013e-07,
+      "loss": 0.7223,
+      "step": 8467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6125573691640296,
+      "learning_rate": 9.913325780064586e-07,
+      "loss": 0.7539,
+      "step": 8468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7445756450530683,
+      "learning_rate": 9.89902957854787e-07,
+      "loss": 0.5746,
+      "step": 8469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6232908198351832,
+      "learning_rate": 9.884743156082445e-07,
+      "loss": 0.6638,
+      "step": 8470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6749833434917487,
+      "learning_rate": 9.870466514218912e-07,
+      "loss": 0.6728,
+      "step": 8471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7949419456953852,
+      "learning_rate": 9.856199654506781e-07,
+      "loss": 0.6282,
+      "step": 8472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5343271623648493,
+      "learning_rate": 9.84194257849448e-07,
+      "loss": 0.6372,
+      "step": 8473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.715316359108728,
+      "learning_rate": 9.827695287729389e-07,
+      "loss": 0.5719,
+      "step": 8474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6786352216456508,
+      "learning_rate": 9.813457783757852e-07,
+      "loss": 0.6733,
+      "step": 8475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6605747453901634,
+      "learning_rate": 9.7992300681251e-07,
+      "loss": 0.7578,
+      "step": 8476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.675631402409404,
+      "learning_rate": 9.78501214237535e-07,
+      "loss": 0.694,
+      "step": 8477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5661955438528077,
+      "learning_rate": 9.770804008051738e-07,
+      "loss": 0.6373,
+      "step": 8478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.583340621840412,
+      "learning_rate": 9.756605666696305e-07,
+      "loss": 0.7857,
+      "step": 8479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.639875934388625,
+      "learning_rate": 9.742417119850078e-07,
+      "loss": 0.6977,
+      "step": 8480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.572932766498734,
+      "learning_rate": 9.72823836905301e-07,
+      "loss": 0.643,
+      "step": 8481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7348617566859073,
+      "learning_rate": 9.714069415843975e-07,
+      "loss": 0.6105,
+      "step": 8482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5129019769957652,
+      "learning_rate": 9.699910261760781e-07,
+      "loss": 0.6587,
+      "step": 8483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7286293367869179,
+      "learning_rate": 9.685760908340215e-07,
+      "loss": 0.6924,
+      "step": 8484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6286863062115362,
+      "learning_rate": 9.671621357117923e-07,
+      "loss": 0.6642,
+      "step": 8485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7886869240578565,
+      "learning_rate": 9.657491609628577e-07,
+      "loss": 0.6126,
+      "step": 8486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6579069465573368,
+      "learning_rate": 9.6433716674057e-07,
+      "loss": 0.6803,
+      "step": 8487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6616465793411628,
+      "learning_rate": 9.629261531981803e-07,
+      "loss": 0.5699,
+      "step": 8488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6520740195369406,
+      "learning_rate": 9.615161204888346e-07,
+      "loss": 0.6986,
+      "step": 8489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7609516169236654,
+      "learning_rate": 9.601070687655667e-07,
+      "loss": 0.6555,
+      "step": 8490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7617317793956033,
+      "learning_rate": 9.586989981813078e-07,
+      "loss": 0.6843,
+      "step": 8491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6235398129125809,
+      "learning_rate": 9.572919088888844e-07,
+      "loss": 0.7243,
+      "step": 8492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.618641132828586,
+      "learning_rate": 9.55885801041011e-07,
+      "loss": 0.675,
+      "step": 8493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7389103985815473,
+      "learning_rate": 9.544806747903013e-07,
+      "loss": 0.7018,
+      "step": 8494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6473771952026806,
+      "learning_rate": 9.530765302892553e-07,
+      "loss": 0.586,
+      "step": 8495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5652406055054269,
+      "learning_rate": 9.516733676902756e-07,
+      "loss": 0.6362,
+      "step": 8496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5478809662387278,
+      "learning_rate": 9.502711871456527e-07,
+      "loss": 0.5799,
+      "step": 8497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5968218731230266,
+      "learning_rate": 9.488699888075681e-07,
+      "loss": 0.695,
+      "step": 8498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9413116133308845,
+      "learning_rate": 9.474697728281058e-07,
+      "loss": 0.6953,
+      "step": 8499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7387764741853589,
+      "learning_rate": 9.460705393592307e-07,
+      "loss": 0.5861,
+      "step": 8500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6879635537094295,
+      "learning_rate": 9.446722885528115e-07,
+      "loss": 0.7975,
+      "step": 8501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7870935928368428,
+      "learning_rate": 9.432750205606079e-07,
+      "loss": 0.6651,
+      "step": 8502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7000098571909354,
+      "learning_rate": 9.418787355342674e-07,
+      "loss": 0.6625,
+      "step": 8503
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6898540784275544,
+      "learning_rate": 9.404834336253366e-07,
+      "loss": 0.6732,
+      "step": 8504
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5420769396071596,
+      "learning_rate": 9.390891149852566e-07,
+      "loss": 0.5964,
+      "step": 8505
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8262952647153974,
+      "learning_rate": 9.376957797653541e-07,
+      "loss": 0.6322,
+      "step": 8506
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6713233513580645,
+      "learning_rate": 9.363034281168581e-07,
+      "loss": 0.7156,
+      "step": 8507
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.558887253750619,
+      "learning_rate": 9.349120601908834e-07,
+      "loss": 0.6006,
+      "step": 8508
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5578407002530088,
+      "learning_rate": 9.335216761384414e-07,
+      "loss": 0.6743,
+      "step": 8509
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4893998244873128,
+      "learning_rate": 9.321322761104401e-07,
+      "loss": 0.5171,
+      "step": 8510
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8732182263707253,
+      "learning_rate": 9.307438602576724e-07,
+      "loss": 0.7097,
+      "step": 8511
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6579487951787695,
+      "learning_rate": 9.293564287308332e-07,
+      "loss": 0.6302,
+      "step": 8512
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5898531234427316,
+      "learning_rate": 9.279699816805032e-07,
+      "loss": 0.7759,
+      "step": 8513
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5373161200871173,
+      "learning_rate": 9.26584519257161e-07,
+      "loss": 0.5888,
+      "step": 8514
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5786124799035222,
+      "learning_rate": 9.252000416111784e-07,
+      "loss": 0.6454,
+      "step": 8515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5839456176787687,
+      "learning_rate": 9.238165488928152e-07,
+      "loss": 0.6728,
+      "step": 8516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7281204784134618,
+      "learning_rate": 9.224340412522325e-07,
+      "loss": 0.6542,
+      "step": 8517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.507704352537572,
+      "learning_rate": 9.210525188394747e-07,
+      "loss": 0.6282,
+      "step": 8518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6918760455004862,
+      "learning_rate": 9.196719818044886e-07,
+      "loss": 0.6208,
+      "step": 8519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7589913626033948,
+      "learning_rate": 9.182924302971086e-07,
+      "loss": 0.6981,
+      "step": 8520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6422584710703088,
+      "learning_rate": 9.169138644670627e-07,
+      "loss": 0.6739,
+      "step": 8521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.767490173430962,
+      "learning_rate": 9.155362844639725e-07,
+      "loss": 0.6822,
+      "step": 8522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8181788312046294,
+      "learning_rate": 9.14159690437355e-07,
+      "loss": 0.7665,
+      "step": 8523
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7480484857115235,
+      "learning_rate": 9.127840825366152e-07,
+      "loss": 0.7395,
+      "step": 8524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8396355344859228,
+      "learning_rate": 9.114094609110569e-07,
+      "loss": 0.582,
+      "step": 8525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6794017948340934,
+      "learning_rate": 9.100358257098707e-07,
+      "loss": 0.8449,
+      "step": 8526
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7382601173910843,
+      "learning_rate": 9.086631770821431e-07,
+      "loss": 0.7845,
+      "step": 8527
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6555141456200846,
+      "learning_rate": 9.07291515176858e-07,
+      "loss": 0.7532,
+      "step": 8528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4531404575168607,
+      "learning_rate": 9.05920840142882e-07,
+      "loss": 0.6653,
+      "step": 8529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9042978793688385,
+      "learning_rate": 9.045511521289862e-07,
+      "loss": 0.6411,
+      "step": 8530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6141730691986127,
+      "learning_rate": 9.031824512838239e-07,
+      "loss": 0.6548,
+      "step": 8531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7912378321838687,
+      "learning_rate": 9.018147377559483e-07,
+      "loss": 0.6434,
+      "step": 8532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6969882014294067,
+      "learning_rate": 9.004480116938041e-07,
+      "loss": 0.6282,
+      "step": 8533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7507718131371228,
+      "learning_rate": 8.99082273245726e-07,
+      "loss": 0.65,
+      "step": 8534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7382108953727158,
+      "learning_rate": 8.977175225599466e-07,
+      "loss": 0.7068,
+      "step": 8535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8785085612766732,
+      "learning_rate": 8.963537597845839e-07,
+      "loss": 0.6955,
+      "step": 8536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9876005931416723,
+      "learning_rate": 8.949909850676564e-07,
+      "loss": 0.796,
+      "step": 8537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8335946031497297,
+      "learning_rate": 8.936291985570722e-07,
+      "loss": 0.5972,
+      "step": 8538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6787740376123934,
+      "learning_rate": 8.922684004006299e-07,
+      "loss": 0.7053,
+      "step": 8539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7260937405636285,
+      "learning_rate": 8.909085907460224e-07,
+      "loss": 0.7055,
+      "step": 8540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8156357333218092,
+      "learning_rate": 8.895497697408395e-07,
+      "loss": 0.7014,
+      "step": 8541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8502892872932764,
+      "learning_rate": 8.881919375325565e-07,
+      "loss": 0.7066,
+      "step": 8542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7520094569059954,
+      "learning_rate": 8.868350942685467e-07,
+      "loss": 0.7422,
+      "step": 8543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6762712922095546,
+      "learning_rate": 8.854792400960721e-07,
+      "loss": 0.6672,
+      "step": 8544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7009066805680277,
+      "learning_rate": 8.841243751622908e-07,
+      "loss": 0.7102,
+      "step": 8545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7598393690465033,
+      "learning_rate": 8.827704996142539e-07,
+      "loss": 0.6883,
+      "step": 8546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6953232913217238,
+      "learning_rate": 8.814176135989006e-07,
+      "loss": 0.6376,
+      "step": 8547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7558576220517113,
+      "learning_rate": 8.800657172630678e-07,
+      "loss": 0.7404,
+      "step": 8548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8020573268309257,
+      "learning_rate": 8.787148107534793e-07,
+      "loss": 0.6439,
+      "step": 8549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7623781627660229,
+      "learning_rate": 8.773648942167578e-07,
+      "loss": 0.7491,
+      "step": 8550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5599459156928088,
+      "learning_rate": 8.760159677994174e-07,
+      "loss": 0.6676,
+      "step": 8551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7111379677657992,
+      "learning_rate": 8.746680316478573e-07,
+      "loss": 0.7308,
+      "step": 8552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7050018071892619,
+      "learning_rate": 8.733210859083807e-07,
+      "loss": 0.7254,
+      "step": 8553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6601874396647305,
+      "learning_rate": 8.719751307271739e-07,
+      "loss": 0.6913,
+      "step": 8554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7903128696098336,
+      "learning_rate": 8.706301662503191e-07,
+      "loss": 0.7505,
+      "step": 8555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6630933552740055,
+      "learning_rate": 8.69286192623795e-07,
+      "loss": 0.6685,
+      "step": 8556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8305512837202105,
+      "learning_rate": 8.679432099934637e-07,
+      "loss": 0.7627,
+      "step": 8557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.644433408519988,
+      "learning_rate": 8.666012185050876e-07,
+      "loss": 0.7553,
+      "step": 8558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7837745416315152,
+      "learning_rate": 8.652602183043213e-07,
+      "loss": 0.7348,
+      "step": 8559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6394931714912497,
+      "learning_rate": 8.63920209536705e-07,
+      "loss": 0.7273,
+      "step": 8560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.759179568705848,
+      "learning_rate": 8.62581192347679e-07,
+      "loss": 0.8114,
+      "step": 8561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9408825514066173,
+      "learning_rate": 8.612431668825705e-07,
+      "loss": 0.7462,
+      "step": 8562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6359865231405288,
+      "learning_rate": 8.59906133286601e-07,
+      "loss": 0.6535,
+      "step": 8563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7203917526452943,
+      "learning_rate": 8.585700917048879e-07,
+      "loss": 0.6331,
+      "step": 8564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.715284772326796,
+      "learning_rate": 8.572350422824338e-07,
+      "loss": 0.7302,
+      "step": 8565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9761566130879007,
+      "learning_rate": 8.559009851641397e-07,
+      "loss": 0.7234,
+      "step": 8566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.60364359740448,
+      "learning_rate": 8.545679204947954e-07,
+      "loss": 0.6343,
+      "step": 8567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7097803626771462,
+      "learning_rate": 8.532358484190851e-07,
+      "loss": 0.6143,
+      "step": 8568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6386662138382595,
+      "learning_rate": 8.519047690815841e-07,
+      "loss": 0.7445,
+      "step": 8569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.692997496401604,
+      "learning_rate": 8.505746826267602e-07,
+      "loss": 0.7129,
+      "step": 8570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7807662897180458,
+      "learning_rate": 8.492455891989737e-07,
+      "loss": 0.7059,
+      "step": 8571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6263058459356212,
+      "learning_rate": 8.479174889424758e-07,
+      "loss": 0.7066,
+      "step": 8572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7883197031359765,
+      "learning_rate": 8.465903820014121e-07,
+      "loss": 0.7816,
+      "step": 8573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6248322732286755,
+      "learning_rate": 8.452642685198209e-07,
+      "loss": 0.8102,
+      "step": 8574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6366081161488895,
+      "learning_rate": 8.43939148641627e-07,
+      "loss": 0.7074,
+      "step": 8575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6256476717105108,
+      "learning_rate": 8.426150225106544e-07,
+      "loss": 0.7085,
+      "step": 8576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6823461597388514,
+      "learning_rate": 8.412918902706169e-07,
+      "loss": 0.8242,
+      "step": 8577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.599186246228301,
+      "learning_rate": 8.399697520651163e-07,
+      "loss": 0.6813,
+      "step": 8578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0396075459147673,
+      "learning_rate": 8.386486080376532e-07,
+      "loss": 0.736,
+      "step": 8579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8394322681952746,
+      "learning_rate": 8.373284583316176e-07,
+      "loss": 0.6502,
+      "step": 8580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.682872999505341,
+      "learning_rate": 8.360093030902871e-07,
+      "loss": 0.6431,
+      "step": 8581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7358414224590166,
+      "learning_rate": 8.346911424568394e-07,
+      "loss": 0.5806,
+      "step": 8582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7439800889886863,
+      "learning_rate": 8.333739765743399e-07,
+      "loss": 0.7109,
+      "step": 8583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.699991984763425,
+      "learning_rate": 8.320578055857432e-07,
+      "loss": 0.7801,
+      "step": 8584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6099028565865006,
+      "learning_rate": 8.307426296339017e-07,
+      "loss": 0.6203,
+      "step": 8585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6382126778019257,
+      "learning_rate": 8.294284488615578e-07,
+      "loss": 0.6541,
+      "step": 8586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.74805337997645,
+      "learning_rate": 8.281152634113432e-07,
+      "loss": 0.7318,
+      "step": 8587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6270872487229284,
+      "learning_rate": 8.268030734257848e-07,
+      "loss": 0.6896,
+      "step": 8588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5915860465129363,
+      "learning_rate": 8.254918790472999e-07,
+      "loss": 0.6669,
+      "step": 8589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5851282974358303,
+      "learning_rate": 8.241816804181968e-07,
+      "loss": 0.5469,
+      "step": 8590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6259615558647373,
+      "learning_rate": 8.228724776806818e-07,
+      "loss": 0.666,
+      "step": 8591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.501919932252582,
+      "learning_rate": 8.215642709768423e-07,
+      "loss": 0.7237,
+      "step": 8592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5567954460938576,
+      "learning_rate": 8.20257060448667e-07,
+      "loss": 0.6374,
+      "step": 8593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.837719264984136,
+      "learning_rate": 8.189508462380335e-07,
+      "loss": 0.7212,
+      "step": 8594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.61006922062226,
+      "learning_rate": 8.176456284867096e-07,
+      "loss": 0.7692,
+      "step": 8595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5625071262859511,
+      "learning_rate": 8.163414073363562e-07,
+      "loss": 0.7378,
+      "step": 8596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.731467292703643,
+      "learning_rate": 8.150381829285282e-07,
+      "loss": 0.6837,
+      "step": 8597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7203878203226242,
+      "learning_rate": 8.137359554046676e-07,
+      "loss": 0.6837,
+      "step": 8598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7322407343324207,
+      "learning_rate": 8.124347249061115e-07,
+      "loss": 0.5974,
+      "step": 8599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.708328769120278,
+      "learning_rate": 8.111344915740893e-07,
+      "loss": 0.7015,
+      "step": 8600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.492761598293974,
+      "learning_rate": 8.098352555497202e-07,
+      "loss": 0.6527,
+      "step": 8601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4512688790266,
+      "learning_rate": 8.085370169740169e-07,
+      "loss": 0.602,
+      "step": 8602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5763027777487575,
+      "learning_rate": 8.072397759878803e-07,
+      "loss": 0.6192,
+      "step": 8603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7378737574428034,
+      "learning_rate": 8.059435327321074e-07,
+      "loss": 0.7466,
+      "step": 8604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7063685731013865,
+      "learning_rate": 8.046482873473871e-07,
+      "loss": 0.6301,
+      "step": 8605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7740020706333472,
+      "learning_rate": 8.033540399742945e-07,
+      "loss": 0.7094,
+      "step": 8606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5270984069291418,
+      "learning_rate": 8.020607907533017e-07,
+      "loss": 0.6618,
+      "step": 8607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.597389958525877,
+      "learning_rate": 8.0076853982477e-07,
+      "loss": 0.699,
+      "step": 8608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5897788497630603,
+      "learning_rate": 7.994772873289536e-07,
+      "loss": 0.76,
+      "step": 8609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6962932147366152,
+      "learning_rate": 7.981870334059983e-07,
+      "loss": 0.6271,
+      "step": 8610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.667922649893392,
+      "learning_rate": 7.968977781959387e-07,
+      "loss": 0.6334,
+      "step": 8611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5968713655185358,
+      "learning_rate": 7.95609521838705e-07,
+      "loss": 0.6992,
+      "step": 8612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5791815893923733,
+      "learning_rate": 7.943222644741189e-07,
+      "loss": 0.6811,
+      "step": 8613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7572013151258459,
+      "learning_rate": 7.930360062418896e-07,
+      "loss": 0.6329,
+      "step": 8614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6938162903103777,
+      "learning_rate": 7.91750747281621e-07,
+      "loss": 0.7315,
+      "step": 8615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6832955244066146,
+      "learning_rate": 7.904664877328072e-07,
+      "loss": 0.6837,
+      "step": 8616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8860479223720243,
+      "learning_rate": 7.891832277348344e-07,
+      "loss": 0.7744,
+      "step": 8617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7369431625286054,
+      "learning_rate": 7.879009674269845e-07,
+      "loss": 0.7475,
+      "step": 8618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5136845094116347,
+      "learning_rate": 7.866197069484205e-07,
+      "loss": 0.6582,
+      "step": 8619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7750629918797527,
+      "learning_rate": 7.85339446438208e-07,
+      "loss": 0.7294,
+      "step": 8620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.587406446054958,
+      "learning_rate": 7.840601860352947e-07,
+      "loss": 0.6335,
+      "step": 8621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6404575019637098,
+      "learning_rate": 7.827819258785285e-07,
+      "loss": 0.7823,
+      "step": 8622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5917314155579665,
+      "learning_rate": 7.815046661066439e-07,
+      "loss": 0.6375,
+      "step": 8623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8137847228022808,
+      "learning_rate": 7.802284068582655e-07,
+      "loss": 0.707,
+      "step": 8624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6504081878106207,
+      "learning_rate": 7.789531482719148e-07,
+      "loss": 0.6542,
+      "step": 8625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.623803955785065,
+      "learning_rate": 7.776788904859956e-07,
+      "loss": 0.6638,
+      "step": 8626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6243223053097606,
+      "learning_rate": 7.764056336388127e-07,
+      "loss": 0.6989,
+      "step": 8627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.798107487378668,
+      "learning_rate": 7.751333778685588e-07,
+      "loss": 0.814,
+      "step": 8628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6166916424175415,
+      "learning_rate": 7.738621233133148e-07,
+      "loss": 0.6075,
+      "step": 8629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4663527224002506,
+      "learning_rate": 7.725918701110557e-07,
+      "loss": 0.5568,
+      "step": 8630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.78406648879962,
+      "learning_rate": 7.713226183996513e-07,
+      "loss": 0.7568,
+      "step": 8631
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.718185309054279,
+      "learning_rate": 7.700543683168537e-07,
+      "loss": 0.6125,
+      "step": 8632
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6783471847093223,
+      "learning_rate": 7.687871200003172e-07,
+      "loss": 0.5707,
+      "step": 8633
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8218426363881475,
+      "learning_rate": 7.675208735875761e-07,
+      "loss": 0.6897,
+      "step": 8634
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7452221894583109,
+      "learning_rate": 7.662556292160639e-07,
+      "loss": 0.7302,
+      "step": 8635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8217020412534808,
+      "learning_rate": 7.649913870231063e-07,
+      "loss": 0.8073,
+      "step": 8636
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9268201950458284,
+      "learning_rate": 7.637281471459124e-07,
+      "loss": 0.6536,
+      "step": 8637
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7537251047143585,
+      "learning_rate": 7.624659097215903e-07,
+      "loss": 0.7079,
+      "step": 8638
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7704002735144673,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.6428,
+      "step": 8639
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7637132366632193,
+      "learning_rate": 7.599444427794301e-07,
+      "loss": 0.6163,
+      "step": 8640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6060984338985305,
+      "learning_rate": 7.586852135352607e-07,
+      "loss": 0.6644,
+      "step": 8641
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.557104348665908,
+      "learning_rate": 7.574269872912921e-07,
+      "loss": 0.6215,
+      "step": 8642
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5334753071785745,
+      "learning_rate": 7.561697641840882e-07,
+      "loss": 0.7568,
+      "step": 8643
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7956769060024518,
+      "learning_rate": 7.549135443500966e-07,
+      "loss": 0.653,
+      "step": 8644
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6257566675549924,
+      "learning_rate": 7.536583279256638e-07,
+      "loss": 0.712,
+      "step": 8645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7807779192366155,
+      "learning_rate": 7.524041150470241e-07,
+      "loss": 0.7125,
+      "step": 8646
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.595006510159101,
+      "learning_rate": 7.511509058502997e-07,
+      "loss": 0.6043,
+      "step": 8647
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7246952203941068,
+      "learning_rate": 7.498987004715108e-07,
+      "loss": 0.67,
+      "step": 8648
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7245351538870213,
+      "learning_rate": 7.486474990465598e-07,
+      "loss": 0.6304,
+      "step": 8649
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6393893794984247,
+      "learning_rate": 7.47397301711249e-07,
+      "loss": 0.5836,
+      "step": 8650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6812533149537447,
+      "learning_rate": 7.461481086012679e-07,
+      "loss": 0.796,
+      "step": 8651
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6652425440486422,
+      "learning_rate": 7.448999198521934e-07,
+      "loss": 0.5835,
+      "step": 8652
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7745124037983675,
+      "learning_rate": 7.436527355994994e-07,
+      "loss": 0.7556,
+      "step": 8653
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6955759539400153,
+      "learning_rate": 7.424065559785498e-07,
+      "loss": 0.758,
+      "step": 8654
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7262393580932531,
+      "learning_rate": 7.411613811245944e-07,
+      "loss": 0.696,
+      "step": 8655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7838213988096059,
+      "learning_rate": 7.399172111727804e-07,
+      "loss": 0.6724,
+      "step": 8656
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7588142980376191,
+      "learning_rate": 7.386740462581399e-07,
+      "loss": 0.7651,
+      "step": 8657
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5901706489580727,
+      "learning_rate": 7.374318865156005e-07,
+      "loss": 0.6707,
+      "step": 8658
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7232413404704006,
+      "learning_rate": 7.361907320799811e-07,
+      "loss": 0.6283,
+      "step": 8659
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6051860022865811,
+      "learning_rate": 7.349505830859871e-07,
+      "loss": 0.5946,
+      "step": 8660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5686212020624424,
+      "learning_rate": 7.337114396682187e-07,
+      "loss": 0.666,
+      "step": 8661
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7657274710165263,
+      "learning_rate": 7.324733019611641e-07,
+      "loss": 0.6962,
+      "step": 8662
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8283228142017602,
+      "learning_rate": 7.312361700992043e-07,
+      "loss": 0.7268,
+      "step": 8663
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.545472626591203,
+      "learning_rate": 7.300000442166133e-07,
+      "loss": 0.7287,
+      "step": 8664
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5451732294601264,
+      "learning_rate": 7.287649244475481e-07,
+      "loss": 0.6112,
+      "step": 8665
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6576688661857788,
+      "learning_rate": 7.27530810926067e-07,
+      "loss": 0.6033,
+      "step": 8666
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8983431471455487,
+      "learning_rate": 7.262977037861096e-07,
+      "loss": 0.6632,
+      "step": 8667
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.539195718199749,
+      "learning_rate": 7.250656031615111e-07,
+      "loss": 0.6502,
+      "step": 8668
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6100641613515339,
+      "learning_rate": 7.23834509186e-07,
+      "loss": 0.6516,
+      "step": 8669
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.712817735434763,
+      "learning_rate": 7.226044219931883e-07,
+      "loss": 0.6382,
+      "step": 8670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.75885847360641,
+      "learning_rate": 7.213753417165836e-07,
+      "loss": 0.7896,
+      "step": 8671
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7738662827225833,
+      "learning_rate": 7.201472684895872e-07,
+      "loss": 0.7514,
+      "step": 8672
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5742849695052534,
+      "learning_rate": 7.189202024454811e-07,
+      "loss": 0.6635,
+      "step": 8673
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6258377366677337,
+      "learning_rate": 7.176941437174489e-07,
+      "loss": 0.6817,
+      "step": 8674
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6906760253031172,
+      "learning_rate": 7.164690924385576e-07,
+      "loss": 0.7801,
+      "step": 8675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5194349350543774,
+      "learning_rate": 7.152450487417673e-07,
+      "loss": 0.6269,
+      "step": 8676
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6111332506431482,
+      "learning_rate": 7.140220127599318e-07,
+      "loss": 0.5915,
+      "step": 8677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6512927420232688,
+      "learning_rate": 7.127999846257893e-07,
+      "loss": 0.6012,
+      "step": 8678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5815162841761377,
+      "learning_rate": 7.115789644719728e-07,
+      "loss": 0.6751,
+      "step": 8679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5351316643915993,
+      "learning_rate": 7.103589524310051e-07,
+      "loss": 0.608,
+      "step": 8680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7310893799644735,
+      "learning_rate": 7.09139948635299e-07,
+      "loss": 0.7092,
+      "step": 8681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.744228287786194,
+      "learning_rate": 7.079219532171599e-07,
+      "loss": 0.722,
+      "step": 8682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6453074599655961,
+      "learning_rate": 7.067049663087788e-07,
+      "loss": 0.7212,
+      "step": 8683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.721431416393383,
+      "learning_rate": 7.054889880422433e-07,
+      "loss": 0.6608,
+      "step": 8684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.603757868663997,
+      "learning_rate": 7.0427401854953e-07,
+      "loss": 0.717,
+      "step": 8685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9247365371113263,
+      "learning_rate": 7.030600579625014e-07,
+      "loss": 0.6379,
+      "step": 8686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8475447505522269,
+      "learning_rate": 7.018471064129162e-07,
+      "loss": 0.7608,
+      "step": 8687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6599123438996048,
+      "learning_rate": 7.006351640324215e-07,
+      "loss": 0.7324,
+      "step": 8688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5674875844829181,
+      "learning_rate": 6.994242309525523e-07,
+      "loss": 0.7213,
+      "step": 8689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9027885332859753,
+      "learning_rate": 6.982143073047387e-07,
+      "loss": 0.7656,
+      "step": 8690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8030207958156752,
+      "learning_rate": 6.970053932202991e-07,
+      "loss": 0.686,
+      "step": 8691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7440038038627508,
+      "learning_rate": 6.957974888304408e-07,
+      "loss": 0.7075,
+      "step": 8692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9613385767698652,
+      "learning_rate": 6.945905942662646e-07,
+      "loss": 0.7465,
+      "step": 8693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.828627449456084,
+      "learning_rate": 6.933847096587575e-07,
+      "loss": 0.7298,
+      "step": 8694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.552784833683063,
+      "learning_rate": 6.92179835138802e-07,
+      "loss": 0.6045,
+      "step": 8695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7410167562173768,
+      "learning_rate": 6.909759708371678e-07,
+      "loss": 0.6272,
+      "step": 8696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5793295105967549,
+      "learning_rate": 6.897731168845145e-07,
+      "loss": 0.6606,
+      "step": 8697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6323565095980503,
+      "learning_rate": 6.885712734113925e-07,
+      "loss": 0.6883,
+      "step": 8698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.460791374584712,
+      "learning_rate": 6.873704405482473e-07,
+      "loss": 0.6915,
+      "step": 8699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5657541778249906,
+      "learning_rate": 6.861706184254046e-07,
+      "loss": 0.6578,
+      "step": 8700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.754491594995093,
+      "learning_rate": 6.849718071730905e-07,
+      "loss": 0.6067,
+      "step": 8701
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.697738919853421,
+      "learning_rate": 6.837740069214161e-07,
+      "loss": 0.6951,
+      "step": 8702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6302173531920074,
+      "learning_rate": 6.825772178003831e-07,
+      "loss": 0.7389,
+      "step": 8703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9243450934911965,
+      "learning_rate": 6.813814399398855e-07,
+      "loss": 0.745,
+      "step": 8704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7761708127088411,
+      "learning_rate": 6.80186673469706e-07,
+      "loss": 0.7524,
+      "step": 8705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.614586091489057,
+      "learning_rate": 6.789929185195166e-07,
+      "loss": 0.7115,
+      "step": 8706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.676059358685063,
+      "learning_rate": 6.778001752188823e-07,
+      "loss": 0.6826,
+      "step": 8707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8001808657182368,
+      "learning_rate": 6.766084436972564e-07,
+      "loss": 0.6911,
+      "step": 8708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7749660708941688,
+      "learning_rate": 6.75417724083981e-07,
+      "loss": 0.6824,
+      "step": 8709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.716846862284347,
+      "learning_rate": 6.742280165082937e-07,
+      "loss": 0.6694,
+      "step": 8710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6762582029546438,
+      "learning_rate": 6.730393210993147e-07,
+      "loss": 0.6982,
+      "step": 8711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7126512371411937,
+      "learning_rate": 6.718516379860595e-07,
+      "loss": 0.6423,
+      "step": 8712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6213435185517746,
+      "learning_rate": 6.706649672974353e-07,
+      "loss": 0.6024,
+      "step": 8713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7864519428301608,
+      "learning_rate": 6.694793091622331e-07,
+      "loss": 0.7844,
+      "step": 8714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.778521136186733,
+      "learning_rate": 6.682946637091404e-07,
+      "loss": 0.6831,
+      "step": 8715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.54754015717152,
+      "learning_rate": 6.671110310667283e-07,
+      "loss": 0.7726,
+      "step": 8716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6000257477234023,
+      "learning_rate": 6.659284113634634e-07,
+      "loss": 0.6161,
+      "step": 8717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5372503227345617,
+      "learning_rate": 6.647468047277029e-07,
+      "loss": 0.5754,
+      "step": 8718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.704517683947603,
+      "learning_rate": 6.635662112876884e-07,
+      "loss": 0.6452,
+      "step": 8719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6588122662892621,
+      "learning_rate": 6.623866311715576e-07,
+      "loss": 0.7267,
+      "step": 8720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.459853858961382,
+      "learning_rate": 6.612080645073316e-07,
+      "loss": 0.5757,
+      "step": 8721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5936972799654623,
+      "learning_rate": 6.600305114229288e-07,
+      "loss": 0.6597,
+      "step": 8722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8774277055499191,
+      "learning_rate": 6.588539720461551e-07,
+      "loss": 0.7892,
+      "step": 8723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7669374648720841,
+      "learning_rate": 6.576784465047014e-07,
+      "loss": 0.7177,
+      "step": 8724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.560008410701888,
+      "learning_rate": 6.565039349261548e-07,
+      "loss": 0.6919,
+      "step": 8725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7644269109338624,
+      "learning_rate": 6.55330437437991e-07,
+      "loss": 0.6835,
+      "step": 8726
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5372222677673302,
+      "learning_rate": 6.541579541675736e-07,
+      "loss": 0.5947,
+      "step": 8727
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8259938415601342,
+      "learning_rate": 6.529864852421586e-07,
+      "loss": 0.7568,
+      "step": 8728
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4769277739964026,
+      "learning_rate": 6.518160307888877e-07,
+      "loss": 0.5646,
+      "step": 8729
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7987995664627618,
+      "learning_rate": 6.506465909347981e-07,
+      "loss": 0.6142,
+      "step": 8730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.477685080019239,
+      "learning_rate": 6.49478165806815e-07,
+      "loss": 0.5199,
+      "step": 8731
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7147832264792515,
+      "learning_rate": 6.483107555317491e-07,
+      "loss": 0.6818,
+      "step": 8732
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7458082121307719,
+      "learning_rate": 6.471443602363093e-07,
+      "loss": 0.7152,
+      "step": 8733
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.745055777773359,
+      "learning_rate": 6.45978980047085e-07,
+      "loss": 0.5758,
+      "step": 8734
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6966452521157125,
+      "learning_rate": 6.44814615090561e-07,
+      "loss": 0.6355,
+      "step": 8735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.633730183689762,
+      "learning_rate": 6.436512654931138e-07,
+      "loss": 0.6368,
+      "step": 8736
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6072410354623625,
+      "learning_rate": 6.424889313810034e-07,
+      "loss": 0.6961,
+      "step": 8737
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8507298257170395,
+      "learning_rate": 6.413276128803858e-07,
+      "loss": 0.6825,
+      "step": 8738
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.746313217113024,
+      "learning_rate": 6.401673101173012e-07,
+      "loss": 0.7445,
+      "step": 8739
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6390561996209958,
+      "learning_rate": 6.390080232176832e-07,
+      "loss": 0.6135,
+      "step": 8740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5743849846487263,
+      "learning_rate": 6.378497523073568e-07,
+      "loss": 0.6436,
+      "step": 8741
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9545625466793428,
+      "learning_rate": 6.366924975120303e-07,
+      "loss": 0.7543,
+      "step": 8742
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6648504342011303,
+      "learning_rate": 6.355362589573078e-07,
+      "loss": 0.6485,
+      "step": 8743
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.716423230872207,
+      "learning_rate": 6.34381036768682e-07,
+      "loss": 0.6911,
+      "step": 8744
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.65445994435496,
+      "learning_rate": 6.332268310715306e-07,
+      "loss": 0.7274,
+      "step": 8745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6244381998771817,
+      "learning_rate": 6.320736419911289e-07,
+      "loss": 0.5986,
+      "step": 8746
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.893230528199649,
+      "learning_rate": 6.309214696526345e-07,
+      "loss": 0.7471,
+      "step": 8747
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6760771995237522,
+      "learning_rate": 6.297703141810973e-07,
+      "loss": 0.7281,
+      "step": 8748
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.758727507648121,
+      "learning_rate": 6.286201757014609e-07,
+      "loss": 0.673,
+      "step": 8749
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6252006539545014,
+      "learning_rate": 6.274710543385498e-07,
+      "loss": 0.6644,
+      "step": 8750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.711993065300969,
+      "learning_rate": 6.263229502170887e-07,
+      "loss": 0.72,
+      "step": 8751
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.598204332420291,
+      "learning_rate": 6.251758634616811e-07,
+      "loss": 0.6851,
+      "step": 8752
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7099373984607578,
+      "learning_rate": 6.240297941968276e-07,
+      "loss": 0.6626,
+      "step": 8753
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.067578511288023,
+      "learning_rate": 6.228847425469176e-07,
+      "loss": 0.7337,
+      "step": 8754
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6329116330234184,
+      "learning_rate": 6.217407086362259e-07,
+      "loss": 0.5908,
+      "step": 8755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8695458114514938,
+      "learning_rate": 6.205976925889223e-07,
+      "loss": 0.6287,
+      "step": 8756
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6009515502837723,
+      "learning_rate": 6.194556945290597e-07,
+      "loss": 0.5986,
+      "step": 8757
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7133996021572895,
+      "learning_rate": 6.183147145805868e-07,
+      "loss": 0.6671,
+      "step": 8758
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.547406379390001,
+      "learning_rate": 6.171747528673399e-07,
+      "loss": 0.5963,
+      "step": 8759
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8441570176178421,
+      "learning_rate": 6.160358095130414e-07,
+      "loss": 0.7358,
+      "step": 8760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5426579135934184,
+      "learning_rate": 6.148978846413067e-07,
+      "loss": 0.6821,
+      "step": 8761
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5955286906448458,
+      "learning_rate": 6.137609783756415e-07,
+      "loss": 0.703,
+      "step": 8762
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9207348627068204,
+      "learning_rate": 6.12625090839436e-07,
+      "loss": 0.761,
+      "step": 8763
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6145845753468058,
+      "learning_rate": 6.11490222155976e-07,
+      "loss": 0.7207,
+      "step": 8764
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8111292561758374,
+      "learning_rate": 6.103563724484318e-07,
+      "loss": 0.5999,
+      "step": 8765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6732627046727127,
+      "learning_rate": 6.092235418398651e-07,
+      "loss": 0.6734,
+      "step": 8766
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8411285377702618,
+      "learning_rate": 6.080917304532297e-07,
+      "loss": 0.6682,
+      "step": 8767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6387647062242647,
+      "learning_rate": 6.069609384113617e-07,
+      "loss": 0.629,
+      "step": 8768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6815980212566355,
+      "learning_rate": 6.058311658369954e-07,
+      "loss": 0.6013,
+      "step": 8769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.706867933567596,
+      "learning_rate": 6.047024128527456e-07,
+      "loss": 0.7289,
+      "step": 8770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.716827005321816,
+      "learning_rate": 6.035746795811226e-07,
+      "loss": 0.6986,
+      "step": 8771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5820734195546344,
+      "learning_rate": 6.024479661445271e-07,
+      "loss": 0.6539,
+      "step": 8772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6433261887082913,
+      "learning_rate": 6.013222726652412e-07,
+      "loss": 0.5254,
+      "step": 8773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6064092532647984,
+      "learning_rate": 6.001975992654452e-07,
+      "loss": 0.5876,
+      "step": 8774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6371089009253044,
+      "learning_rate": 5.990739460672024e-07,
+      "loss": 0.6335,
+      "step": 8775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9318912881798678,
+      "learning_rate": 5.979513131924686e-07,
+      "loss": 0.7516,
+      "step": 8776
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.696039269204613,
+      "learning_rate": 5.968297007630897e-07,
+      "loss": 0.7741,
+      "step": 8777
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.704697818598336,
+      "learning_rate": 5.95709108900796e-07,
+      "loss": 0.6288,
+      "step": 8778
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6282649929871835,
+      "learning_rate": 5.945895377272115e-07,
+      "loss": 0.6602,
+      "step": 8779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7299319787649439,
+      "learning_rate": 5.9347098736385e-07,
+      "loss": 0.6479,
+      "step": 8780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5855476532183754,
+      "learning_rate": 5.9235345793211e-07,
+      "loss": 0.7551,
+      "step": 8781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9476339944739738,
+      "learning_rate": 5.912369495532844e-07,
+      "loss": 0.8062,
+      "step": 8782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8287390115049857,
+      "learning_rate": 5.901214623485507e-07,
+      "loss": 0.7113,
+      "step": 8783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7171646753440621,
+      "learning_rate": 5.890069964389766e-07,
+      "loss": 0.8326,
+      "step": 8784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6001861328056501,
+      "learning_rate": 5.878935519455243e-07,
+      "loss": 0.6403,
+      "step": 8785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8358706633985116,
+      "learning_rate": 5.867811289890357e-07,
+      "loss": 0.6297,
+      "step": 8786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8183021473531173,
+      "learning_rate": 5.856697276902512e-07,
+      "loss": 0.7184,
+      "step": 8787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7328689608493777,
+      "learning_rate": 5.845593481697931e-07,
+      "loss": 0.6518,
+      "step": 8788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7449931725599808,
+      "learning_rate": 5.834499905481761e-07,
+      "loss": 0.6504,
+      "step": 8789
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7985757893626173,
+      "learning_rate": 5.823416549458061e-07,
+      "loss": 0.768,
+      "step": 8790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7058853899108344,
+      "learning_rate": 5.812343414829725e-07,
+      "loss": 0.6564,
+      "step": 8791
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5406487918403375,
+      "learning_rate": 5.801280502798578e-07,
+      "loss": 0.6524,
+      "step": 8792
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7917898832172519,
+      "learning_rate": 5.790227814565342e-07,
+      "loss": 0.6613,
+      "step": 8793
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6094745785554516,
+      "learning_rate": 5.779185351329586e-07,
+      "loss": 0.677,
+      "step": 8794
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6648861762187483,
+      "learning_rate": 5.76815311428982e-07,
+      "loss": 0.7195,
+      "step": 8795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7563619296436717,
+      "learning_rate": 5.75713110464341e-07,
+      "loss": 0.7305,
+      "step": 8796
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7070913640267633,
+      "learning_rate": 5.746119323586619e-07,
+      "loss": 0.5819,
+      "step": 8797
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.791324973188892,
+      "learning_rate": 5.735117772314625e-07,
+      "loss": 0.6723,
+      "step": 8798
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5459587913165356,
+      "learning_rate": 5.724126452021439e-07,
+      "loss": 0.6511,
+      "step": 8799
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7139177229387026,
+      "learning_rate": 5.713145363900019e-07,
+      "loss": 0.6557,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6248645438784817,
+      "learning_rate": 5.702174509142211e-07,
+      "loss": 0.6548,
+      "step": 8801
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5587540887616815,
+      "learning_rate": 5.691213888938684e-07,
+      "loss": 0.7533,
+      "step": 8802
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8162267140969848,
+      "learning_rate": 5.680263504479067e-07,
+      "loss": 0.691,
+      "step": 8803
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9258082565468186,
+      "learning_rate": 5.669323356951873e-07,
+      "loss": 0.7071,
+      "step": 8804
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8151940950777297,
+      "learning_rate": 5.658393447544441e-07,
+      "loss": 0.7587,
+      "step": 8805
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5083937395603395,
+      "learning_rate": 5.647473777443057e-07,
+      "loss": 0.6167,
+      "step": 8806
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6689479666754305,
+      "learning_rate": 5.636564347832907e-07,
+      "loss": 0.645,
+      "step": 8807
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6817830151070023,
+      "learning_rate": 5.625665159897997e-07,
+      "loss": 0.6561,
+      "step": 8808
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5860333611721988,
+      "learning_rate": 5.614776214821293e-07,
+      "loss": 0.6598,
+      "step": 8809
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5677057151523508,
+      "learning_rate": 5.603897513784628e-07,
+      "loss": 0.6577,
+      "step": 8810
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7430558459350933,
+      "learning_rate": 5.593029057968668e-07,
+      "loss": 0.7343,
+      "step": 8811
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8275938824697127,
+      "learning_rate": 5.582170848553059e-07,
+      "loss": 0.7346,
+      "step": 8812
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6236573604172517,
+      "learning_rate": 5.571322886716279e-07,
+      "loss": 0.6967,
+      "step": 8813
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8136108917306695,
+      "learning_rate": 5.560485173635699e-07,
+      "loss": 0.6979,
+      "step": 8814
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5996349472778744,
+      "learning_rate": 5.549657710487588e-07,
+      "loss": 0.6484,
+      "step": 8815
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9571357824539959,
+      "learning_rate": 5.538840498447084e-07,
+      "loss": 0.7165,
+      "step": 8816
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7955011318416891,
+      "learning_rate": 5.528033538688227e-07,
+      "loss": 0.723,
+      "step": 8817
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6455737570632334,
+      "learning_rate": 5.517236832383976e-07,
+      "loss": 0.6117,
+      "step": 8818
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5258600663929245,
+      "learning_rate": 5.506450380706108e-07,
+      "loss": 0.5365,
+      "step": 8819
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5290508706361885,
+      "learning_rate": 5.495674184825339e-07,
+      "loss": 0.6308,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4450450071362761,
+      "learning_rate": 5.484908245911269e-07,
+      "loss": 0.5544,
+      "step": 8821
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.861407949886194,
+      "learning_rate": 5.47415256513234e-07,
+      "loss": 0.7452,
+      "step": 8822
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6397086283679179,
+      "learning_rate": 5.463407143655941e-07,
+      "loss": 0.6864,
+      "step": 8823
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6945030783026827,
+      "learning_rate": 5.452671982648306e-07,
+      "loss": 0.5986,
+      "step": 8824
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1115877118850843,
+      "learning_rate": 5.441947083274568e-07,
+      "loss": 0.6233,
+      "step": 8825
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6672585918040428,
+      "learning_rate": 5.431232446698764e-07,
+      "loss": 0.68,
+      "step": 8826
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7129379266769043,
+      "learning_rate": 5.420528074083775e-07,
+      "loss": 0.7318,
+      "step": 8827
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7850490370440566,
+      "learning_rate": 5.409833966591416e-07,
+      "loss": 0.7114,
+      "step": 8828
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6887809896724186,
+      "learning_rate": 5.399150125382346e-07,
+      "loss": 0.619,
+      "step": 8829
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7080850987518374,
+      "learning_rate": 5.38847655161614e-07,
+      "loss": 0.7035,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5918728979806458,
+      "learning_rate": 5.377813246451258e-07,
+      "loss": 0.6376,
+      "step": 8831
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6662792606891546,
+      "learning_rate": 5.367160211044997e-07,
+      "loss": 0.5893,
+      "step": 8832
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.052849741595967,
+      "learning_rate": 5.356517446553632e-07,
+      "loss": 0.6952,
+      "step": 8833
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8600307295170615,
+      "learning_rate": 5.345884954132219e-07,
+      "loss": 0.8211,
+      "step": 8834
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7060491542482932,
+      "learning_rate": 5.335262734934777e-07,
+      "loss": 0.7927,
+      "step": 8835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6548610817413254,
+      "learning_rate": 5.324650790114183e-07,
+      "loss": 0.6175,
+      "step": 8836
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5534172736814402,
+      "learning_rate": 5.314049120822173e-07,
+      "loss": 0.6578,
+      "step": 8837
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6066570013867365,
+      "learning_rate": 5.303457728209405e-07,
+      "loss": 0.6216,
+      "step": 8838
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7314963661007452,
+      "learning_rate": 5.292876613425435e-07,
+      "loss": 0.5863,
+      "step": 8839
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8081504341128891,
+      "learning_rate": 5.282305777618635e-07,
+      "loss": 0.6855,
+      "step": 8840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6112349432475361,
+      "learning_rate": 5.271745221936342e-07,
+      "loss": 0.7426,
+      "step": 8841
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8055474774673523,
+      "learning_rate": 5.261194947524706e-07,
+      "loss": 0.7428,
+      "step": 8842
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5931208961495853,
+      "learning_rate": 5.2506549555288e-07,
+      "loss": 0.6494,
+      "step": 8843
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.640245601508315,
+      "learning_rate": 5.240125247092609e-07,
+      "loss": 0.6547,
+      "step": 8844
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6570113616081257,
+      "learning_rate": 5.22960582335893e-07,
+      "loss": 0.6057,
+      "step": 8845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7830829445931253,
+      "learning_rate": 5.219096685469505e-07,
+      "loss": 0.6648,
+      "step": 8846
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5182028049341616,
+      "learning_rate": 5.20859783456491e-07,
+      "loss": 0.6532,
+      "step": 8847
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6825405089912515,
+      "learning_rate": 5.198109271784657e-07,
+      "loss": 0.7685,
+      "step": 8848
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5108201298654471,
+      "learning_rate": 5.187630998267112e-07,
+      "loss": 0.6859,
+      "step": 8849
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7284137246283864,
+      "learning_rate": 5.177163015149522e-07,
+      "loss": 0.7895,
+      "step": 8850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6011817453650048,
+      "learning_rate": 5.166705323568022e-07,
+      "loss": 0.7195,
+      "step": 8851
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6879633198914614,
+      "learning_rate": 5.156257924657626e-07,
+      "loss": 0.6305,
+      "step": 8852
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5509576011661719,
+      "learning_rate": 5.145820819552239e-07,
+      "loss": 0.5572,
+      "step": 8853
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6281892283913746,
+      "learning_rate": 5.135394009384665e-07,
+      "loss": 0.7006,
+      "step": 8854
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5648948142189403,
+      "learning_rate": 5.124977495286543e-07,
+      "loss": 0.634,
+      "step": 8855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6985997006874525,
+      "learning_rate": 5.114571278388436e-07,
+      "loss": 0.662,
+      "step": 8856
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7954802063819606,
+      "learning_rate": 5.104175359819785e-07,
+      "loss": 0.7451,
+      "step": 8857
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6763239305455557,
+      "learning_rate": 5.093789740708877e-07,
+      "loss": 0.6913,
+      "step": 8858
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7869358727317133,
+      "learning_rate": 5.083414422182942e-07,
+      "loss": 0.7111,
+      "step": 8859
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7085254944971797,
+      "learning_rate": 5.073049405368025e-07,
+      "loss": 0.8868,
+      "step": 8860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7339717725846475,
+      "learning_rate": 5.062694691389114e-07,
+      "loss": 0.7482,
+      "step": 8861
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5826376971671454,
+      "learning_rate": 5.052350281370045e-07,
+      "loss": 0.7272,
+      "step": 8862
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7025735469207472,
+      "learning_rate": 5.042016176433529e-07,
+      "loss": 0.7245,
+      "step": 8863
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7507929274931837,
+      "learning_rate": 5.031692377701192e-07,
+      "loss": 0.7461,
+      "step": 8864
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7509772240735764,
+      "learning_rate": 5.021378886293493e-07,
+      "loss": 0.7327,
+      "step": 8865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5796503675823719,
+      "learning_rate": 5.011075703329816e-07,
+      "loss": 0.6357,
+      "step": 8866
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6355729150637843,
+      "learning_rate": 5.00078282992843e-07,
+      "loss": 0.6922,
+      "step": 8867
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9593713687066243,
+      "learning_rate": 4.990500267206422e-07,
+      "loss": 0.6745,
+      "step": 8868
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8007290348212777,
+      "learning_rate": 4.980228016279853e-07,
+      "loss": 0.6228,
+      "step": 8869
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5854190455395145,
+      "learning_rate": 4.969966078263578e-07,
+      "loss": 0.6157,
+      "step": 8870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7029363611122843,
+      "learning_rate": 4.95971445427137e-07,
+      "loss": 0.6451,
+      "step": 8871
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7041474249843223,
+      "learning_rate": 4.949473145415917e-07,
+      "loss": 0.6162,
+      "step": 8872
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9193299614643426,
+      "learning_rate": 4.939242152808709e-07,
+      "loss": 0.6908,
+      "step": 8873
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6604203951924335,
+      "learning_rate": 4.929021477560181e-07,
+      "loss": 0.6437,
+      "step": 8874
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7807366861294014,
+      "learning_rate": 4.918811120779655e-07,
+      "loss": 0.7067,
+      "step": 8875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8913639048565587,
+      "learning_rate": 4.908611083575254e-07,
+      "loss": 0.6881,
+      "step": 8876
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.605533038984582,
+      "learning_rate": 4.898421367054063e-07,
+      "loss": 0.6292,
+      "step": 8877
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7706508305498252,
+      "learning_rate": 4.888241972322005e-07,
+      "loss": 0.7253,
+      "step": 8878
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7098943647143408,
+      "learning_rate": 4.878072900483899e-07,
+      "loss": 0.6648,
+      "step": 8879
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7194533016750646,
+      "learning_rate": 4.86791415264346e-07,
+      "loss": 0.7055,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6214507345058031,
+      "learning_rate": 4.857765729903219e-07,
+      "loss": 0.6689,
+      "step": 8881
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5346747157802176,
+      "learning_rate": 4.84762763336466e-07,
+      "loss": 0.641,
+      "step": 8882
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5087485363749233,
+      "learning_rate": 4.837499864128104e-07,
+      "loss": 0.573,
+      "step": 8883
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7344990112796763,
+      "learning_rate": 4.827382423292748e-07,
+      "loss": 0.723,
+      "step": 8884
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7580447093681373,
+      "learning_rate": 4.817275311956715e-07,
+      "loss": 0.6838,
+      "step": 8885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7700545438137756,
+      "learning_rate": 4.80717853121695e-07,
+      "loss": 0.8107,
+      "step": 8886
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.608306797865548,
+      "learning_rate": 4.797092082169308e-07,
+      "loss": 0.6479,
+      "step": 8887
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8590731878633593,
+      "learning_rate": 4.787015965908504e-07,
+      "loss": 0.7478,
+      "step": 8888
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9107457527445748,
+      "learning_rate": 4.776950183528161e-07,
+      "loss": 0.6514,
+      "step": 8889
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7168042445704972,
+      "learning_rate": 4.766894736120753e-07,
+      "loss": 0.6907,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8278719101055414,
+      "learning_rate": 4.7568496247776373e-07,
+      "loss": 0.7917,
+      "step": 8891
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.530010529849326,
+      "learning_rate": 4.746814850589054e-07,
+      "loss": 0.6842,
+      "step": 8892
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.488161536929116,
+      "learning_rate": 4.7367904146441435e-07,
+      "loss": 0.6651,
+      "step": 8893
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8409654711792904,
+      "learning_rate": 4.7267763180308676e-07,
+      "loss": 0.719,
+      "step": 8894
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.883523084424903,
+      "learning_rate": 4.716772561836136e-07,
+      "loss": 0.7707,
+      "step": 8895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.771972446027783,
+      "learning_rate": 4.706779147145657e-07,
+      "loss": 0.6725,
+      "step": 8896
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6792519467501912,
+      "learning_rate": 4.6967960750440855e-07,
+      "loss": 0.6237,
+      "step": 8897
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7675836119266148,
+      "learning_rate": 4.6868233466149327e-07,
+      "loss": 0.7123,
+      "step": 8898
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6269745798391906,
+      "learning_rate": 4.676860962940577e-07,
+      "loss": 0.6876,
+      "step": 8899
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7769618813041788,
+      "learning_rate": 4.6669089251022757e-07,
+      "loss": 0.6378,
+      "step": 8900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5590363379953227,
+      "learning_rate": 4.6569672341801544e-07,
+      "loss": 0.7152,
+      "step": 8901
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8800667696896778,
+      "learning_rate": 4.647035891253249e-07,
+      "loss": 0.68,
+      "step": 8902
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7554247388764033,
+      "learning_rate": 4.637114897399453e-07,
+      "loss": 0.7968,
+      "step": 8903
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5907387453462725,
+      "learning_rate": 4.6272042536955164e-07,
+      "loss": 0.7354,
+      "step": 8904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.718431267879274,
+      "learning_rate": 4.61730396121709e-07,
+      "loss": 0.6396,
+      "step": 8905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7295315815770955,
+      "learning_rate": 4.607414021038714e-07,
+      "loss": 0.695,
+      "step": 8906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.66657712292234,
+      "learning_rate": 4.5975344342337525e-07,
+      "loss": 0.646,
+      "step": 8907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.620845024767798,
+      "learning_rate": 4.5876652018745136e-07,
+      "loss": 0.6812,
+      "step": 8908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5818359888527045,
+      "learning_rate": 4.5778063250321415e-07,
+      "loss": 0.6347,
+      "step": 8909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.575993128204435,
+      "learning_rate": 4.567957804776646e-07,
+      "loss": 0.6201,
+      "step": 8910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.71544634015368,
+      "learning_rate": 4.558119642176939e-07,
+      "loss": 0.6866,
+      "step": 8911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8006115963235978,
+      "learning_rate": 4.548291838300811e-07,
+      "loss": 0.6697,
+      "step": 8912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6565075655767947,
+      "learning_rate": 4.5384743942148977e-07,
+      "loss": 0.7009,
+      "step": 8913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7720099961704778,
+      "learning_rate": 4.5286673109847357e-07,
+      "loss": 0.6587,
+      "step": 8914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7665613437823307,
+      "learning_rate": 4.518870589674751e-07,
+      "loss": 0.6843,
+      "step": 8915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6085851494760013,
+      "learning_rate": 4.509084231348182e-07,
+      "loss": 0.6389,
+      "step": 8916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5926002358110811,
+      "learning_rate": 4.499308237067213e-07,
+      "loss": 0.7338,
+      "step": 8917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6111636451724387,
+      "learning_rate": 4.489542607892894e-07,
+      "loss": 0.6652,
+      "step": 8918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7833895836947011,
+      "learning_rate": 4.479787344885078e-07,
+      "loss": 0.8026,
+      "step": 8919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6879517925619654,
+      "learning_rate": 4.470042449102596e-07,
+      "loss": 0.6402,
+      "step": 8920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8778262970244906,
+      "learning_rate": 4.4603079216030797e-07,
+      "loss": 0.7691,
+      "step": 8921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7471342930414981,
+      "learning_rate": 4.4505837634430616e-07,
+      "loss": 0.7868,
+      "step": 8922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4882165634147697,
+      "learning_rate": 4.440869975677964e-07,
+      "loss": 0.6551,
+      "step": 8923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7413866953660853,
+      "learning_rate": 4.4311665593620323e-07,
+      "loss": 0.6163,
+      "step": 8924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5929686827582712,
+      "learning_rate": 4.421473515548447e-07,
+      "loss": 0.6359,
+      "step": 8925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7485565665955185,
+      "learning_rate": 4.411790845289243e-07,
+      "loss": 0.6752,
+      "step": 8926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5022991327522344,
+      "learning_rate": 4.4021185496353036e-07,
+      "loss": 0.537,
+      "step": 8927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7332070844888663,
+      "learning_rate": 4.392456629636399e-07,
+      "loss": 0.6884,
+      "step": 8928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6941139000307985,
+      "learning_rate": 4.382805086341213e-07,
+      "loss": 0.6324,
+      "step": 8929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8277559933369407,
+      "learning_rate": 4.373163920797241e-07,
+      "loss": 0.6778,
+      "step": 8930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8101863523173403,
+      "learning_rate": 4.36353313405089e-07,
+      "loss": 0.7158,
+      "step": 8931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6692541425579466,
+      "learning_rate": 4.353912727147425e-07,
+      "loss": 0.7753,
+      "step": 8932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.718180299453628,
+      "learning_rate": 4.344302701130998e-07,
+      "loss": 0.6736,
+      "step": 8933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7976225939029056,
+      "learning_rate": 4.3347030570446314e-07,
+      "loss": 0.6929,
+      "step": 8934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0948448179186894,
+      "learning_rate": 4.3251137959302023e-07,
+      "loss": 0.7218,
+      "step": 8935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6384943101690133,
+      "learning_rate": 4.3155349188284903e-07,
+      "loss": 0.7612,
+      "step": 8936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7067810498067775,
+      "learning_rate": 4.305966426779118e-07,
+      "loss": 0.6568,
+      "step": 8937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7004994744578605,
+      "learning_rate": 4.2964083208206e-07,
+      "loss": 0.5903,
+      "step": 8938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7022825552040735,
+      "learning_rate": 4.286860601990328e-07,
+      "loss": 0.7522,
+      "step": 8939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6869400366291563,
+      "learning_rate": 4.2773232713245515e-07,
+      "loss": 0.7511,
+      "step": 8940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6267627208860855,
+      "learning_rate": 4.267796329858398e-07,
+      "loss": 0.6479,
+      "step": 8941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5098530524214537,
+      "learning_rate": 4.2582797786258735e-07,
+      "loss": 0.5881,
+      "step": 8942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.678415811161424,
+      "learning_rate": 4.2487736186598297e-07,
+      "loss": 0.7102,
+      "step": 8943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.811580603149009,
+      "learning_rate": 4.2392778509920516e-07,
+      "loss": 0.6838,
+      "step": 8944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9324127880015471,
+      "learning_rate": 4.229792476653116e-07,
+      "loss": 0.8057,
+      "step": 8945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8912088902745035,
+      "learning_rate": 4.2203174966725324e-07,
+      "loss": 0.7762,
+      "step": 8946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.607190207534778,
+      "learning_rate": 4.210852912078656e-07,
+      "loss": 0.643,
+      "step": 8947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7417460029699203,
+      "learning_rate": 4.2013987238987197e-07,
+      "loss": 0.7045,
+      "step": 8948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7337059018397725,
+      "learning_rate": 4.1919549331588374e-07,
+      "loss": 0.599,
+      "step": 8949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7286472884952369,
+      "learning_rate": 4.182521540883966e-07,
+      "loss": 0.7168,
+      "step": 8950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6656075871975395,
+      "learning_rate": 4.1730985480979645e-07,
+      "loss": 0.6924,
+      "step": 8951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6302672025258802,
+      "learning_rate": 4.1636859558235707e-07,
+      "loss": 0.7471,
+      "step": 8952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6921406858909696,
+      "learning_rate": 4.154283765082334e-07,
+      "loss": 0.7204,
+      "step": 8953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7806522406686651,
+      "learning_rate": 4.144891976894749e-07,
+      "loss": 0.6436,
+      "step": 8954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.872644394026765,
+      "learning_rate": 4.135510592280112e-07,
+      "loss": 0.7537,
+      "step": 8955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6822983087884402,
+      "learning_rate": 4.126139612256663e-07,
+      "loss": 0.632,
+      "step": 8956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6924687648307566,
+      "learning_rate": 4.1167790378414674e-07,
+      "loss": 0.601,
+      "step": 8957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6927001381944409,
+      "learning_rate": 4.107428870050445e-07,
+      "loss": 0.7585,
+      "step": 8958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6511589387598713,
+      "learning_rate": 4.09808910989844e-07,
+      "loss": 0.6367,
+      "step": 8959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7125663757115257,
+      "learning_rate": 4.088759758399108e-07,
+      "loss": 0.681,
+      "step": 8960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8104751149832377,
+      "learning_rate": 4.079440816565028e-07,
+      "loss": 0.6205,
+      "step": 8961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6243580652214884,
+      "learning_rate": 4.070132285407624e-07,
+      "loss": 0.7469,
+      "step": 8962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8825980210289652,
+      "learning_rate": 4.0608341659371774e-07,
+      "loss": 0.7058,
+      "step": 8963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6623960622591962,
+      "learning_rate": 4.0515464591628696e-07,
+      "loss": 0.6874,
+      "step": 8964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5681784036517072,
+      "learning_rate": 4.042269166092716e-07,
+      "loss": 0.8016,
+      "step": 8965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7266043370859399,
+      "learning_rate": 4.0330022877336337e-07,
+      "loss": 0.6505,
+      "step": 8966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.781958601565994,
+      "learning_rate": 4.023745825091407e-07,
+      "loss": 0.628,
+      "step": 8967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7222770015039668,
+      "learning_rate": 4.0144997791706664e-07,
+      "loss": 0.6065,
+      "step": 8968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7610167368399874,
+      "learning_rate": 4.005264150974919e-07,
+      "loss": 0.7912,
+      "step": 8969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6324699798367595,
+      "learning_rate": 3.9960389415065746e-07,
+      "loss": 0.6925,
+      "step": 8970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0978893687631777,
+      "learning_rate": 3.986824151766855e-07,
+      "loss": 0.708,
+      "step": 8971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5194927759176355,
+      "learning_rate": 3.977619782755915e-07,
+      "loss": 0.5378,
+      "step": 8972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7378334005633167,
+      "learning_rate": 3.968425835472711e-07,
+      "loss": 0.6148,
+      "step": 8973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8144345071554742,
+      "learning_rate": 3.959242310915112e-07,
+      "loss": 0.7466,
+      "step": 8974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8078036734219984,
+      "learning_rate": 3.9500692100798656e-07,
+      "loss": 0.7714,
+      "step": 8975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.617085545757556,
+      "learning_rate": 3.9409065339625407e-07,
+      "loss": 0.6702,
+      "step": 8976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7290694019525905,
+      "learning_rate": 3.9317542835576317e-07,
+      "loss": 0.707,
+      "step": 8977
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6922559342798456,
+      "learning_rate": 3.9226124598584323e-07,
+      "loss": 0.6726,
+      "step": 8978
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9071775216199967,
+      "learning_rate": 3.913481063857183e-07,
+      "loss": 0.6511,
+      "step": 8979
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5998712830431994,
+      "learning_rate": 3.904360096544935e-07,
+      "loss": 0.7038,
+      "step": 8980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6427375601725964,
+      "learning_rate": 3.895249558911629e-07,
+      "loss": 0.672,
+      "step": 8981
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6755330097592833,
+      "learning_rate": 3.8861494519460865e-07,
+      "loss": 0.6397,
+      "step": 8982
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6342444584998612,
+      "learning_rate": 3.87705977663595e-07,
+      "loss": 0.6957,
+      "step": 8983
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.619564578301826,
+      "learning_rate": 3.867980533967786e-07,
+      "loss": 0.6582,
+      "step": 8984
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5857180219289881,
+      "learning_rate": 3.858911724927006e-07,
+      "loss": 0.6964,
+      "step": 8985
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6850104647723942,
+      "learning_rate": 3.849853350497879e-07,
+      "loss": 0.7092,
+      "step": 8986
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6564782792292456,
+      "learning_rate": 3.84080541166354e-07,
+      "loss": 0.7242,
+      "step": 8987
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7271970757460506,
+      "learning_rate": 3.831767909406026e-07,
+      "loss": 0.6899,
+      "step": 8988
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8320539782625036,
+      "learning_rate": 3.8227408447061853e-07,
+      "loss": 0.774,
+      "step": 8989
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8335983391878803,
+      "learning_rate": 3.813724218543802e-07,
+      "loss": 0.7207,
+      "step": 8990
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6308164696229996,
+      "learning_rate": 3.8047180318974474e-07,
+      "loss": 0.7297,
+      "step": 8991
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3703166795086241,
+      "learning_rate": 3.7957222857446297e-07,
+      "loss": 0.6287,
+      "step": 8992
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7678726313798627,
+      "learning_rate": 3.7867369810617114e-07,
+      "loss": 0.7225,
+      "step": 8993
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.561412053470494,
+      "learning_rate": 3.777762118823869e-07,
+      "loss": 0.6146,
+      "step": 8994
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7741039057431165,
+      "learning_rate": 3.768797700005211e-07,
+      "loss": 0.7123,
+      "step": 8995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.809479864639746,
+      "learning_rate": 3.75984372557866e-07,
+      "loss": 0.7771,
+      "step": 8996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.029375389380463,
+      "learning_rate": 3.7509001965160494e-07,
+      "loss": 0.6924,
+      "step": 8997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7141203015363913,
+      "learning_rate": 3.74196711378807e-07,
+      "loss": 0.6498,
+      "step": 8998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5722400717490779,
+      "learning_rate": 3.733044478364234e-07,
+      "loss": 0.6627,
+      "step": 8999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7598986440797793,
+      "learning_rate": 3.72413229121299e-07,
+      "loss": 0.6553,
+      "step": 9000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7653352445711585,
+      "learning_rate": 3.715230553301585e-07,
+      "loss": 0.689,
+      "step": 9001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.887705746303693,
+      "learning_rate": 3.706339265596182e-07,
+      "loss": 0.8099,
+      "step": 9002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5972384367430816,
+      "learning_rate": 3.6974584290617955e-07,
+      "loss": 0.6685,
+      "step": 9003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7070696333151447,
+      "learning_rate": 3.688588044662289e-07,
+      "loss": 0.6961,
+      "step": 9004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7616232130572633,
+      "learning_rate": 3.6797281133603926e-07,
+      "loss": 0.6333,
+      "step": 9005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7008064665959346,
+      "learning_rate": 3.6708786361177586e-07,
+      "loss": 0.725,
+      "step": 9006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6544849763586182,
+      "learning_rate": 3.662039613894808e-07,
+      "loss": 0.6224,
+      "step": 9007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6644316689410115,
+      "learning_rate": 3.6532110476509064e-07,
+      "loss": 0.6801,
+      "step": 9008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7592494890729364,
+      "learning_rate": 3.644392938344254e-07,
+      "loss": 0.7281,
+      "step": 9009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7409770219731049,
+      "learning_rate": 3.6355852869318976e-07,
+      "loss": 0.6187,
+      "step": 9010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7447307456725696,
+      "learning_rate": 3.6267880943698153e-07,
+      "loss": 0.6459,
+      "step": 9011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.76726525973141,
+      "learning_rate": 3.6180013616127554e-07,
+      "loss": 0.7181,
+      "step": 9012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.74741714153782,
+      "learning_rate": 3.6092250896143986e-07,
+      "loss": 0.6133,
+      "step": 9013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8751732333083337,
+      "learning_rate": 3.6004592793272954e-07,
+      "loss": 0.7615,
+      "step": 9014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6830093484207342,
+      "learning_rate": 3.5917039317028057e-07,
+      "loss": 0.7213,
+      "step": 9015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7216518180084097,
+      "learning_rate": 3.5829590476911925e-07,
+      "loss": 0.8031,
+      "step": 9016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6066940758025394,
+      "learning_rate": 3.574224628241596e-07,
+      "loss": 0.6596,
+      "step": 9017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.741718527853461,
+      "learning_rate": 3.5655006743019695e-07,
+      "loss": 0.6479,
+      "step": 9018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.956259263388941,
+      "learning_rate": 3.556787186819177e-07,
+      "loss": 0.7758,
+      "step": 9019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.664427510137072,
+      "learning_rate": 3.548084166738952e-07,
+      "loss": 0.682,
+      "step": 9020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.393762831262044,
+      "learning_rate": 3.5393916150058274e-07,
+      "loss": 0.5206,
+      "step": 9021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6615508318160366,
+      "learning_rate": 3.5307095325632814e-07,
+      "loss": 0.7007,
+      "step": 9022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8205337937904904,
+      "learning_rate": 3.522037920353605e-07,
+      "loss": 0.669,
+      "step": 9023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8779304425650445,
+      "learning_rate": 3.5133767793179676e-07,
+      "loss": 0.7525,
+      "step": 9024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7115456730628904,
+      "learning_rate": 3.5047261103963837e-07,
+      "loss": 0.6273,
+      "step": 9025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7026022015340625,
+      "learning_rate": 3.49608591452778e-07,
+      "loss": 0.7278,
+      "step": 9026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5533481880995001,
+      "learning_rate": 3.4874561926498964e-07,
+      "loss": 0.52,
+      "step": 9027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6861445365830685,
+      "learning_rate": 3.478836945699349e-07,
+      "loss": 0.706,
+      "step": 9028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7246204553515043,
+      "learning_rate": 3.470228174611634e-07,
+      "loss": 0.6582,
+      "step": 9029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7259912400155208,
+      "learning_rate": 3.461629880321082e-07,
+      "loss": 0.6964,
+      "step": 9030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5258391955713864,
+      "learning_rate": 3.4530420637609365e-07,
+      "loss": 0.6223,
+      "step": 9031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9501077686189958,
+      "learning_rate": 3.444464725863228e-07,
+      "loss": 0.7221,
+      "step": 9032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.519065257951722,
+      "learning_rate": 3.435897867558924e-07,
+      "loss": 0.7602,
+      "step": 9033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6559610091291315,
+      "learning_rate": 3.4273414897778133e-07,
+      "loss": 0.6338,
+      "step": 9034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8860713373495857,
+      "learning_rate": 3.418795593448554e-07,
+      "loss": 0.6843,
+      "step": 9035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7613812992796087,
+      "learning_rate": 3.4102601794986813e-07,
+      "loss": 0.6712,
+      "step": 9036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6267258861120946,
+      "learning_rate": 3.401735248854554e-07,
+      "loss": 0.6966,
+      "step": 9037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7168945312821067,
+      "learning_rate": 3.3932208024414435e-07,
+      "loss": 0.8096,
+      "step": 9038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8824721722015318,
+      "learning_rate": 3.3847168411834666e-07,
+      "loss": 0.7043,
+      "step": 9039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.644816582797017,
+      "learning_rate": 3.3762233660035724e-07,
+      "loss": 0.7229,
+      "step": 9040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6122426967789936,
+      "learning_rate": 3.367740377823603e-07,
+      "loss": 0.6559,
+      "step": 9041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7948017004198267,
+      "learning_rate": 3.3592678775642653e-07,
+      "loss": 0.7649,
+      "step": 9042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6036517898197797,
+      "learning_rate": 3.35080586614509e-07,
+      "loss": 0.7244,
+      "step": 9043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5588599713340119,
+      "learning_rate": 3.342354344484533e-07,
+      "loss": 0.6241,
+      "step": 9044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7085589108728771,
+      "learning_rate": 3.333913313499848e-07,
+      "loss": 0.642,
+      "step": 9045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.756713623597841,
+      "learning_rate": 3.3254827741071806e-07,
+      "loss": 0.6681,
+      "step": 9046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4899810166224117,
+      "learning_rate": 3.3170627272215427e-07,
+      "loss": 0.6863,
+      "step": 9047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.531364479212317,
+      "learning_rate": 3.308653173756793e-07,
+      "loss": 0.755,
+      "step": 9048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5168149460058475,
+      "learning_rate": 3.300254114625656e-07,
+      "loss": 0.7084,
+      "step": 9049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7459672256196828,
+      "learning_rate": 3.2918655507397144e-07,
+      "loss": 0.7533,
+      "step": 9050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7158933208599039,
+      "learning_rate": 3.283487483009429e-07,
+      "loss": 0.6339,
+      "step": 9051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7430776876204692,
+      "learning_rate": 3.2751199123441046e-07,
+      "loss": 0.6434,
+      "step": 9052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5704904697330333,
+      "learning_rate": 3.2667628396518936e-07,
+      "loss": 0.5842,
+      "step": 9053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7800244667694278,
+      "learning_rate": 3.258416265839848e-07,
+      "loss": 0.7669,
+      "step": 9054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7638354518588866,
+      "learning_rate": 3.2500801918138425e-07,
+      "loss": 0.5612,
+      "step": 9055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8344341984251336,
+      "learning_rate": 3.241754618478632e-07,
+      "loss": 0.647,
+      "step": 9056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7039427641608778,
+      "learning_rate": 3.2334395467378266e-07,
+      "loss": 0.6021,
+      "step": 9057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7322214223361194,
+      "learning_rate": 3.2251349774939046e-07,
+      "loss": 0.6456,
+      "step": 9058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7517206466744195,
+      "learning_rate": 3.216840911648178e-07,
+      "loss": 0.7309,
+      "step": 9059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6778998604568132,
+      "learning_rate": 3.2085573501008717e-07,
+      "loss": 0.6758,
+      "step": 9060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8758051183890794,
+      "learning_rate": 3.200284293750999e-07,
+      "loss": 0.7864,
+      "step": 9061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5772792146727868,
+      "learning_rate": 3.1920217434964985e-07,
+      "loss": 0.6732,
+      "step": 9062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8914308924735,
+      "learning_rate": 3.1837697002341293e-07,
+      "loss": 0.6585,
+      "step": 9063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7453450043856464,
+      "learning_rate": 3.1755281648595093e-07,
+      "loss": 0.7913,
+      "step": 9064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3858743411899503,
+      "learning_rate": 3.1672971382671556e-07,
+      "loss": 0.6196,
+      "step": 9065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8338183072323526,
+      "learning_rate": 3.159076621350399e-07,
+      "loss": 0.7064,
+      "step": 9066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.685855771910675,
+      "learning_rate": 3.1508666150014575e-07,
+      "loss": 0.6349,
+      "step": 9067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6489397211084076,
+      "learning_rate": 3.142667120111387e-07,
+      "loss": 0.6519,
+      "step": 9068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.49196056325252,
+      "learning_rate": 3.134478137570118e-07,
+      "loss": 0.6841,
+      "step": 9069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6392600812772116,
+      "learning_rate": 3.126299668266453e-07,
+      "loss": 0.6858,
+      "step": 9070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6652442805073417,
+      "learning_rate": 3.1181317130880127e-07,
+      "loss": 0.6811,
+      "step": 9071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.708651031293616,
+      "learning_rate": 3.1099742729213235e-07,
+      "loss": 0.71,
+      "step": 9072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.754886721993718,
+      "learning_rate": 3.101827348651731e-07,
+      "loss": 0.6408,
+      "step": 9073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6962840997056352,
+      "learning_rate": 3.093690941163452e-07,
+      "loss": 0.6917,
+      "step": 9074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5700005927580967,
+      "learning_rate": 3.08556505133959e-07,
+      "loss": 0.6271,
+      "step": 9075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.581948249429852,
+      "learning_rate": 3.077449680062061e-07,
+      "loss": 0.6254,
+      "step": 9076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.51273905759815,
+      "learning_rate": 3.069344828211662e-07,
+      "loss": 0.6443,
+      "step": 9077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7028985474900407,
+      "learning_rate": 3.061250496668078e-07,
+      "loss": 0.749,
+      "step": 9078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6677452775571646,
+      "learning_rate": 3.053166686309783e-07,
+      "loss": 0.6592,
+      "step": 9079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6062980603348134,
+      "learning_rate": 3.0450933980141763e-07,
+      "loss": 0.6491,
+      "step": 9080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5340881022151536,
+      "learning_rate": 3.0370306326574673e-07,
+      "loss": 0.6637,
+      "step": 9081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6672173189261768,
+      "learning_rate": 3.028978391114745e-07,
+      "loss": 0.7864,
+      "step": 9082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.736912952197448,
+      "learning_rate": 3.020936674259989e-07,
+      "loss": 0.6965,
+      "step": 9083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.688420218707243,
+      "learning_rate": 3.012905482965944e-07,
+      "loss": 0.6519,
+      "step": 9084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.735607676551278,
+      "learning_rate": 3.0048848181043256e-07,
+      "loss": 0.673,
+      "step": 9085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7040912601415914,
+      "learning_rate": 2.996874680545603e-07,
+      "loss": 0.7363,
+      "step": 9086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6373307067637413,
+      "learning_rate": 2.9888750711591805e-07,
+      "loss": 0.6322,
+      "step": 9087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8062957317799482,
+      "learning_rate": 2.980885990813298e-07,
+      "loss": 0.6328,
+      "step": 9088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5429622083207397,
+      "learning_rate": 2.972907440375017e-07,
+      "loss": 0.6703,
+      "step": 9089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7175453590035683,
+      "learning_rate": 2.964939420710311e-07,
+      "loss": 0.7612,
+      "step": 9090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6746756006078176,
+      "learning_rate": 2.956981932683967e-07,
+      "loss": 0.697,
+      "step": 9091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7654789376597413,
+      "learning_rate": 2.949034977159648e-07,
+      "loss": 0.6529,
+      "step": 9092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5521819010013271,
+      "learning_rate": 2.941098554999877e-07,
+      "loss": 0.6326,
+      "step": 9093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7965320659002997,
+      "learning_rate": 2.933172667066031e-07,
+      "loss": 0.6972,
+      "step": 9094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5800734066412943,
+      "learning_rate": 2.9252573142183327e-07,
+      "loss": 0.6148,
+      "step": 9095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8429652754631338,
+      "learning_rate": 2.917352497315873e-07,
+      "loss": 0.6747,
+      "step": 9096
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6399270165705966,
+      "learning_rate": 2.9094582172165876e-07,
+      "loss": 0.6913,
+      "step": 9097
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6281374340831094,
+      "learning_rate": 2.9015744747773024e-07,
+      "loss": 0.7202,
+      "step": 9098
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.552728970873856,
+      "learning_rate": 2.893701270853655e-07,
+      "loss": 0.752,
+      "step": 9099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.65165872718086,
+      "learning_rate": 2.885838606300151e-07,
+      "loss": 0.6869,
+      "step": 9100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7091754810620192,
+      "learning_rate": 2.8779864819701853e-07,
+      "loss": 0.6345,
+      "step": 9101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6686871905612022,
+      "learning_rate": 2.8701448987159654e-07,
+      "loss": 0.5509,
+      "step": 9102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8173373699523554,
+      "learning_rate": 2.8623138573885767e-07,
+      "loss": 0.7899,
+      "step": 9103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7692576744780697,
+      "learning_rate": 2.854493358837951e-07,
+      "loss": 0.6359,
+      "step": 9104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8505447600940077,
+      "learning_rate": 2.8466834039128754e-07,
+      "loss": 0.6465,
+      "step": 9105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7751451897404646,
+      "learning_rate": 2.838883993461028e-07,
+      "loss": 0.7206,
+      "step": 9106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8723784794316785,
+      "learning_rate": 2.831095128328876e-07,
+      "loss": 0.6643,
+      "step": 9107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6200263587394934,
+      "learning_rate": 2.823316809361809e-07,
+      "loss": 0.5683,
+      "step": 9108
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.702563720865403,
+      "learning_rate": 2.8155490374040196e-07,
+      "loss": 0.7154,
+      "step": 9109
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7451961706558843,
+      "learning_rate": 2.807791813298588e-07,
+      "loss": 0.7522,
+      "step": 9110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6920928919156772,
+      "learning_rate": 2.8000451378874525e-07,
+      "loss": 0.6407,
+      "step": 9111
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7605926944594885,
+      "learning_rate": 2.7923090120113626e-07,
+      "loss": 0.6672,
+      "step": 9112
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5573327560649162,
+      "learning_rate": 2.7845834365099913e-07,
+      "loss": 0.6749,
+      "step": 9113
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.616019278808511,
+      "learning_rate": 2.7768684122217893e-07,
+      "loss": 0.6522,
+      "step": 9114
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7863348392332858,
+      "learning_rate": 2.769163939984121e-07,
+      "loss": 0.6702,
+      "step": 9115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9669771441858406,
+      "learning_rate": 2.7614700206332056e-07,
+      "loss": 0.7218,
+      "step": 9116
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6853909847735868,
+      "learning_rate": 2.7537866550040647e-07,
+      "loss": 0.5787,
+      "step": 9117
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7771673876183924,
+      "learning_rate": 2.74611384393062e-07,
+      "loss": 0.698,
+      "step": 9118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5863480135223418,
+      "learning_rate": 2.7384515882456386e-07,
+      "loss": 0.685,
+      "step": 9119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6153180520052017,
+      "learning_rate": 2.730799888780744e-07,
+      "loss": 0.6658,
+      "step": 9120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5779497539387755,
+      "learning_rate": 2.7231587463663945e-07,
+      "loss": 0.627,
+      "step": 9121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8041035551184352,
+      "learning_rate": 2.7155281618319265e-07,
+      "loss": 0.7443,
+      "step": 9122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8765336984772065,
+      "learning_rate": 2.7079081360055106e-07,
+      "loss": 0.6708,
+      "step": 9123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5160141983328415,
+      "learning_rate": 2.7002986697141966e-07,
+      "loss": 0.5983,
+      "step": 9124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.854303661617946,
+      "learning_rate": 2.692699763783868e-07,
+      "loss": 0.7064,
+      "step": 9125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5622552568194452,
+      "learning_rate": 2.685111419039255e-07,
+      "loss": 0.6858,
+      "step": 9126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5987403896005212,
+      "learning_rate": 2.677533636303964e-07,
+      "loss": 0.6825,
+      "step": 9127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6486100461672395,
+      "learning_rate": 2.669966416400449e-07,
+      "loss": 0.7009,
+      "step": 9128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8947303431333133,
+      "learning_rate": 2.6624097601499976e-07,
+      "loss": 0.7161,
+      "step": 9129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6099671906456852,
+      "learning_rate": 2.654863668372787e-07,
+      "loss": 0.5869,
+      "step": 9130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8616912431349213,
+      "learning_rate": 2.6473281418878173e-07,
+      "loss": 0.6781,
+      "step": 9131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.87333267111817,
+      "learning_rate": 2.6398031815129454e-07,
+      "loss": 0.8606,
+      "step": 9132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7026026627353354,
+      "learning_rate": 2.6322887880649074e-07,
+      "loss": 0.5823,
+      "step": 9133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.663095785623857,
+      "learning_rate": 2.6247849623592504e-07,
+      "loss": 0.613,
+      "step": 9134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7113272331280165,
+      "learning_rate": 2.617291705210412e-07,
+      "loss": 0.691,
+      "step": 9135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8262977928445028,
+      "learning_rate": 2.6098090174316636e-07,
+      "loss": 0.6946,
+      "step": 9136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7582436591373753,
+      "learning_rate": 2.602336899835134e-07,
+      "loss": 0.6886,
+      "step": 9137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8419965747676532,
+      "learning_rate": 2.594875353231796e-07,
+      "loss": 0.7765,
+      "step": 9138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.622396277431157,
+      "learning_rate": 2.5874243784315025e-07,
+      "loss": 0.5949,
+      "step": 9139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.635845730948259,
+      "learning_rate": 2.579983976242928e-07,
+      "loss": 0.6032,
+      "step": 9140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9015091477785202,
+      "learning_rate": 2.572554147473616e-07,
+      "loss": 0.7317,
+      "step": 9141
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6575673802244377,
+      "learning_rate": 2.565134892929955e-07,
+      "loss": 0.6806,
+      "step": 9142
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5589129509589466,
+      "learning_rate": 2.5577262134171887e-07,
+      "loss": 0.666,
+      "step": 9143
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7090507291266002,
+      "learning_rate": 2.5503281097394194e-07,
+      "loss": 0.6619,
+      "step": 9144
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7464178687329157,
+      "learning_rate": 2.5429405826995934e-07,
+      "loss": 0.688,
+      "step": 9145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.585521708580421,
+      "learning_rate": 2.5355636330994915e-07,
+      "loss": 0.6253,
+      "step": 9146
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6957372162455264,
+      "learning_rate": 2.5281972617398064e-07,
+      "loss": 0.7472,
+      "step": 9147
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7136949583543848,
+      "learning_rate": 2.52084146942001e-07,
+      "loss": 0.6604,
+      "step": 9148
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5735246095230109,
+      "learning_rate": 2.5134962569384746e-07,
+      "loss": 0.7197,
+      "step": 9149
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4968907551306718,
+      "learning_rate": 2.506161625092396e-07,
+      "loss": 0.5578,
+      "step": 9150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9557448471366423,
+      "learning_rate": 2.498837574677837e-07,
+      "loss": 0.7901,
+      "step": 9151
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6133938766958,
+      "learning_rate": 2.491524106489718e-07,
+      "loss": 0.792,
+      "step": 9152
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.734088575672457,
+      "learning_rate": 2.484221221321792e-07,
+      "loss": 0.7127,
+      "step": 9153
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7922240857189475,
+      "learning_rate": 2.47692891996667e-07,
+      "loss": 0.6896,
+      "step": 9154
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5268419657742744,
+      "learning_rate": 2.4696472032158305e-07,
+      "loss": 0.5813,
+      "step": 9155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7239929918768928,
+      "learning_rate": 2.462376071859585e-07,
+      "loss": 0.7521,
+      "step": 9156
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6846588000318006,
+      "learning_rate": 2.4551155266871017e-07,
+      "loss": 0.6519,
+      "step": 9157
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8150620464602065,
+      "learning_rate": 2.447865568486385e-07,
+      "loss": 0.6218,
+      "step": 9158
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7133394743606516,
+      "learning_rate": 2.440626198044327e-07,
+      "loss": 0.6594,
+      "step": 9159
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9240325579831237,
+      "learning_rate": 2.4333974161466324e-07,
+      "loss": 0.7677,
+      "step": 9160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8672205907680561,
+      "learning_rate": 2.4261792235778737e-07,
+      "loss": 0.6923,
+      "step": 9161
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5890577955916911,
+      "learning_rate": 2.418971621121491e-07,
+      "loss": 0.6073,
+      "step": 9162
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6869253072456738,
+      "learning_rate": 2.411774609559725e-07,
+      "loss": 0.6317,
+      "step": 9163
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6117386715023836,
+      "learning_rate": 2.404588189673718e-07,
+      "loss": 0.5953,
+      "step": 9164
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.695397700532162,
+      "learning_rate": 2.3974123622434566e-07,
+      "loss": 0.6871,
+      "step": 9165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6048962661185335,
+      "learning_rate": 2.390247128047729e-07,
+      "loss": 0.6849,
+      "step": 9166
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4696291515897013,
+      "learning_rate": 2.383092487864247e-07,
+      "loss": 0.6443,
+      "step": 9167
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4987936697717255,
+      "learning_rate": 2.3759484424695113e-07,
+      "loss": 0.6706,
+      "step": 9168
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7643556140864138,
+      "learning_rate": 2.3688149926389015e-07,
+      "loss": 0.6497,
+      "step": 9169
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6598835699626207,
+      "learning_rate": 2.361692139146643e-07,
+      "loss": 0.6873,
+      "step": 9170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6054352567438275,
+      "learning_rate": 2.354579882765806e-07,
+      "loss": 0.6001,
+      "step": 9171
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8758251422175432,
+      "learning_rate": 2.3474782242683292e-07,
+      "loss": 0.7149,
+      "step": 9172
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1173561572795223,
+      "learning_rate": 2.3403871644249731e-07,
+      "loss": 0.6189,
+      "step": 9173
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6994143128015073,
+      "learning_rate": 2.3333067040053558e-07,
+      "loss": 0.6513,
+      "step": 9174
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6467310548387322,
+      "learning_rate": 2.3262368437779736e-07,
+      "loss": 0.7023,
+      "step": 9175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.601730453794561,
+      "learning_rate": 2.3191775845101238e-07,
+      "loss": 0.6781,
+      "step": 9176
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6270212271280904,
+      "learning_rate": 2.3121289269679937e-07,
+      "loss": 0.5984,
+      "step": 9177
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6570884566322033,
+      "learning_rate": 2.3050908719166155e-07,
+      "loss": 0.6897,
+      "step": 9178
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6162197492295876,
+      "learning_rate": 2.2980634201198227e-07,
+      "loss": 0.6794,
+      "step": 9179
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8218868665569699,
+      "learning_rate": 2.2910465723403717e-07,
+      "loss": 0.7722,
+      "step": 9180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6672965841302891,
+      "learning_rate": 2.2840403293398095e-07,
+      "loss": 0.7197,
+      "step": 9181
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7340015665776833,
+      "learning_rate": 2.2770446918785716e-07,
+      "loss": 0.7679,
+      "step": 9182
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6038653974037864,
+      "learning_rate": 2.2700596607159175e-07,
+      "loss": 0.6896,
+      "step": 9183
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7365201637290837,
+      "learning_rate": 2.263085236609952e-07,
+      "loss": 0.6372,
+      "step": 9184
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7596204753893094,
+      "learning_rate": 2.2561214203176583e-07,
+      "loss": 0.7103,
+      "step": 9185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7182996150070722,
+      "learning_rate": 2.249168212594832e-07,
+      "loss": 0.6525,
+      "step": 9186
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8044017196554798,
+      "learning_rate": 2.2422256141961473e-07,
+      "loss": 0.7075,
+      "step": 9187
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6296108621493695,
+      "learning_rate": 2.2352936258751235e-07,
+      "loss": 0.5933,
+      "step": 9188
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.587229110833575,
+      "learning_rate": 2.228372248384092e-07,
+      "loss": 0.6692,
+      "step": 9189
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4784135454097325,
+      "learning_rate": 2.2214614824742853e-07,
+      "loss": 0.6358,
+      "step": 9190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7090710335488055,
+      "learning_rate": 2.214561328895748e-07,
+      "loss": 0.6261,
+      "step": 9191
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.842598291770892,
+      "learning_rate": 2.2076717883973808e-07,
+      "loss": 0.6602,
+      "step": 9192
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7598264214121744,
+      "learning_rate": 2.2007928617269414e-07,
+      "loss": 0.6436,
+      "step": 9193
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8035073777880144,
+      "learning_rate": 2.1939245496310324e-07,
+      "loss": 0.7111,
+      "step": 9194
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6420986225961434,
+      "learning_rate": 2.187066852855091e-07,
+      "loss": 0.6878,
+      "step": 9195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6271321324358616,
+      "learning_rate": 2.1802197721434215e-07,
+      "loss": 0.6896,
+      "step": 9196
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7339393507898984,
+      "learning_rate": 2.1733833082391632e-07,
+      "loss": 0.7423,
+      "step": 9197
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8347943762378955,
+      "learning_rate": 2.166557461884322e-07,
+      "loss": 0.7663,
+      "step": 9198
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6972038780825494,
+      "learning_rate": 2.1597422338197172e-07,
+      "loss": 0.7218,
+      "step": 9199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.034996875430456,
+      "learning_rate": 2.1529376247850342e-07,
+      "loss": 0.6533,
+      "step": 9200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.603699345889999,
+      "learning_rate": 2.146143635518827e-07,
+      "loss": 0.7444,
+      "step": 9201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.763026468203593,
+      "learning_rate": 2.139360266758461e-07,
+      "loss": 0.6964,
+      "step": 9202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5337279377060986,
+      "learning_rate": 2.13258751924017e-07,
+      "loss": 0.7364,
+      "step": 9203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6065746501014624,
+      "learning_rate": 2.1258253936990213e-07,
+      "loss": 0.7114,
+      "step": 9204
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6610368388773546,
+      "learning_rate": 2.1190738908689502e-07,
+      "loss": 0.63,
+      "step": 9205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6533510918146666,
+      "learning_rate": 2.1123330114827256e-07,
+      "loss": 0.6356,
+      "step": 9206
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7435137130120424,
+      "learning_rate": 2.1056027562719517e-07,
+      "loss": 0.7129,
+      "step": 9207
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5918102940992866,
+      "learning_rate": 2.0988831259670994e-07,
+      "loss": 0.5952,
+      "step": 9208
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7703162917156607,
+      "learning_rate": 2.0921741212974967e-07,
+      "loss": 0.6956,
+      "step": 9209
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.725491826595931,
+      "learning_rate": 2.085475742991272e-07,
+      "loss": 0.6831,
+      "step": 9210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.522459509380362,
+      "learning_rate": 2.078787991775455e-07,
+      "loss": 0.5566,
+      "step": 9211
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5164115294750733,
+      "learning_rate": 2.072110868375876e-07,
+      "loss": 0.629,
+      "step": 9212
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6799430968173943,
+      "learning_rate": 2.0654443735172447e-07,
+      "loss": 0.6195,
+      "step": 9213
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7529038428627008,
+      "learning_rate": 2.058788507923104e-07,
+      "loss": 0.6232,
+      "step": 9214
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6082086073174993,
+      "learning_rate": 2.052143272315843e-07,
+      "loss": 0.6703,
+      "step": 9215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6880699939080532,
+      "learning_rate": 2.0455086674166957e-07,
+      "loss": 0.7222,
+      "step": 9216
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6500289904255732,
+      "learning_rate": 2.038884693945742e-07,
+      "loss": 0.6659,
+      "step": 9217
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5552756948520066,
+      "learning_rate": 2.0322713526219062e-07,
+      "loss": 0.6413,
+      "step": 9218
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6993894982632298,
+      "learning_rate": 2.0256686441629814e-07,
+      "loss": 0.694,
+      "step": 9219
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7077427956672064,
+      "learning_rate": 2.019076569285583e-07,
+      "loss": 0.6796,
+      "step": 9220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.700456720601353,
+      "learning_rate": 2.0124951287051718e-07,
+      "loss": 0.6839,
+      "step": 9221
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7073823170278,
+      "learning_rate": 2.0059243231360437e-07,
+      "loss": 0.63,
+      "step": 9222
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7442798621212803,
+      "learning_rate": 1.999364153291383e-07,
+      "loss": 0.6994,
+      "step": 9223
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.003575890582916,
+      "learning_rate": 1.9928146198831987e-07,
+      "loss": 0.7964,
+      "step": 9224
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7730842780328422,
+      "learning_rate": 1.9862757236223108e-07,
+      "loss": 0.7864,
+      "step": 9225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5536572444139543,
+      "learning_rate": 1.979747465218429e-07,
+      "loss": 0.6897,
+      "step": 9226
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7481819975287163,
+      "learning_rate": 1.9732298453801092e-07,
+      "loss": 0.6845,
+      "step": 9227
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.744509227646625,
+      "learning_rate": 1.9667228648147074e-07,
+      "loss": 0.6998,
+      "step": 9228
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7519756322022366,
+      "learning_rate": 1.9602265242284813e-07,
+      "loss": 0.7248,
+      "step": 9229
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6632754313037552,
+      "learning_rate": 1.9537408243264888e-07,
+      "loss": 0.673,
+      "step": 9230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8540429133686656,
+      "learning_rate": 1.947265765812656e-07,
+      "loss": 0.7019,
+      "step": 9231
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7358934404834045,
+      "learning_rate": 1.9408013493897537e-07,
+      "loss": 0.7499,
+      "step": 9232
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.579191613274508,
+      "learning_rate": 1.9343475757593987e-07,
+      "loss": 0.6206,
+      "step": 9233
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1684428344582893,
+      "learning_rate": 1.9279044456220307e-07,
+      "loss": 0.6809,
+      "step": 9234
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7471847525721256,
+      "learning_rate": 1.921471959676957e-07,
+      "loss": 0.6492,
+      "step": 9235
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.679965055034842,
+      "learning_rate": 1.91505011862233e-07,
+      "loss": 0.6013,
+      "step": 9236
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6388236682172188,
+      "learning_rate": 1.908638923155126e-07,
+      "loss": 0.6785,
+      "step": 9237
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5508458601759716,
+      "learning_rate": 1.9022383739712104e-07,
+      "loss": 0.7022,
+      "step": 9238
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7548631919890791,
+      "learning_rate": 1.895848471765227e-07,
+      "loss": 0.733,
+      "step": 9239
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.663254244472624,
+      "learning_rate": 1.8894692172307106e-07,
+      "loss": 0.6702,
+      "step": 9240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6462632236988877,
+      "learning_rate": 1.8831006110600404e-07,
+      "loss": 0.8247,
+      "step": 9241
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6436328578247896,
+      "learning_rate": 1.8767426539444188e-07,
+      "loss": 0.6637,
+      "step": 9242
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.692688762963928,
+      "learning_rate": 1.8703953465739055e-07,
+      "loss": 0.735,
+      "step": 9243
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6500218020690514,
+      "learning_rate": 1.8640586896374157e-07,
+      "loss": 0.5977,
+      "step": 9244
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7399479592499723,
+      "learning_rate": 1.857732683822666e-07,
+      "loss": 0.6792,
+      "step": 9245
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8599038601472067,
+      "learning_rate": 1.851417329816263e-07,
+      "loss": 0.6901,
+      "step": 9246
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.572795620263803,
+      "learning_rate": 1.8451126283036358e-07,
+      "loss": 0.5979,
+      "step": 9247
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7503420322332646,
+      "learning_rate": 1.8388185799690705e-07,
+      "loss": 0.7422,
+      "step": 9248
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8567826988069571,
+      "learning_rate": 1.8325351854956652e-07,
+      "loss": 0.7061,
+      "step": 9249
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7429096529080754,
+      "learning_rate": 1.8262624455654077e-07,
+      "loss": 0.65,
+      "step": 9250
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6255268881969265,
+      "learning_rate": 1.8200003608590977e-07,
+      "loss": 0.6084,
+      "step": 9251
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.602696923514729,
+      "learning_rate": 1.8137489320563806e-07,
+      "loss": 0.6819,
+      "step": 9252
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8004520193464986,
+      "learning_rate": 1.807508159835758e-07,
+      "loss": 0.6275,
+      "step": 9253
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8567473176405347,
+      "learning_rate": 1.8012780448745548e-07,
+      "loss": 0.7466,
+      "step": 9254
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7660728132743209,
+      "learning_rate": 1.7950585878489856e-07,
+      "loss": 0.6696,
+      "step": 9255
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6680895285956447,
+      "learning_rate": 1.7888497894340328e-07,
+      "loss": 0.6764,
+      "step": 9256
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8890462461888413,
+      "learning_rate": 1.7826516503036018e-07,
+      "loss": 0.6705,
+      "step": 9257
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6092121708191458,
+      "learning_rate": 1.7764641711303764e-07,
+      "loss": 0.6481,
+      "step": 9258
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.853872643159592,
+      "learning_rate": 1.77028735258592e-07,
+      "loss": 0.6965,
+      "step": 9259
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6994009901835503,
+      "learning_rate": 1.764121195340629e-07,
+      "loss": 0.657,
+      "step": 9260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6002015651261523,
+      "learning_rate": 1.7579657000637464e-07,
+      "loss": 0.6356,
+      "step": 9261
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7882959951825168,
+      "learning_rate": 1.7518208674233595e-07,
+      "loss": 0.8028,
+      "step": 9262
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5715626462526937,
+      "learning_rate": 1.7456866980863795e-07,
+      "loss": 0.5792,
+      "step": 9263
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.606745758826103,
+      "learning_rate": 1.7395631927185853e-07,
+      "loss": 0.6333,
+      "step": 9264
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.588132832958181,
+      "learning_rate": 1.7334503519846002e-07,
+      "loss": 0.7629,
+      "step": 9265
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5870608189041842,
+      "learning_rate": 1.7273481765478383e-07,
+      "loss": 0.6958,
+      "step": 9266
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7650485962558329,
+      "learning_rate": 1.7212566670706366e-07,
+      "loss": 0.7763,
+      "step": 9267
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6304775041889885,
+      "learning_rate": 1.7151758242141102e-07,
+      "loss": 0.8069,
+      "step": 9268
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6757597922240886,
+      "learning_rate": 1.7091056486382428e-07,
+      "loss": 0.7548,
+      "step": 9269
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6025496608955452,
+      "learning_rate": 1.7030461410018628e-07,
+      "loss": 0.6505,
+      "step": 9270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.609210244897395,
+      "learning_rate": 1.696997301962633e-07,
+      "loss": 0.6027,
+      "step": 9271
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.56754740181272,
+      "learning_rate": 1.69095913217705e-07,
+      "loss": 0.6646,
+      "step": 9272
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5314525225926743,
+      "learning_rate": 1.6849316323004794e-07,
+      "loss": 0.5952,
+      "step": 9273
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.704004564930101,
+      "learning_rate": 1.6789148029871083e-07,
+      "loss": 0.7793,
+      "step": 9274
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5067426360899583,
+      "learning_rate": 1.672908644889959e-07,
+      "loss": 0.5874,
+      "step": 9275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7121136462489004,
+      "learning_rate": 1.666913158660921e-07,
+      "loss": 0.6971,
+      "step": 9276
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5541399878664655,
+      "learning_rate": 1.6609283449506853e-07,
+      "loss": 0.6459,
+      "step": 9277
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.651047465860301,
+      "learning_rate": 1.654954204408843e-07,
+      "loss": 0.7469,
+      "step": 9278
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7380456494653596,
+      "learning_rate": 1.6489907376837644e-07,
+      "loss": 0.6762,
+      "step": 9279
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8908809620929607,
+      "learning_rate": 1.643037945422721e-07,
+      "loss": 0.7793,
+      "step": 9280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5530664329085104,
+      "learning_rate": 1.6370958282717619e-07,
+      "loss": 0.716,
+      "step": 9281
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9011341123308512,
+      "learning_rate": 1.6311643868758274e-07,
+      "loss": 0.6896,
+      "step": 9282
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7537647607023992,
+      "learning_rate": 1.6252436218786915e-07,
+      "loss": 0.712,
+      "step": 9283
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.756264117673827,
+      "learning_rate": 1.619333533922951e-07,
+      "loss": 0.7068,
+      "step": 9284
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7453694164411129,
+      "learning_rate": 1.6134341236500373e-07,
+      "loss": 0.625,
+      "step": 9285
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.612092023051016,
+      "learning_rate": 1.6075453917002827e-07,
+      "loss": 0.68,
+      "step": 9286
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4284971486950722,
+      "learning_rate": 1.6016673387127645e-07,
+      "loss": 0.6504,
+      "step": 9287
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8336605844552831,
+      "learning_rate": 1.595799965325495e-07,
+      "loss": 0.7225,
+      "step": 9288
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7925815124293898,
+      "learning_rate": 1.589943272175265e-07,
+      "loss": 0.6753,
+      "step": 9289
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.833749887623125,
+      "learning_rate": 1.5840972598977212e-07,
+      "loss": 0.6691,
+      "step": 9290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7636615851990618,
+      "learning_rate": 1.5782619291273894e-07,
+      "loss": 0.6865,
+      "step": 9291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7029992810691348,
+      "learning_rate": 1.572437280497563e-07,
+      "loss": 0.7755,
+      "step": 9292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6787600559848652,
+      "learning_rate": 1.5666233146404474e-07,
+      "loss": 0.5814,
+      "step": 9293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6399835606911737,
+      "learning_rate": 1.5608200321870382e-07,
+      "loss": 0.5949,
+      "step": 9294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9466123957984622,
+      "learning_rate": 1.5550274337671868e-07,
+      "loss": 0.6788,
+      "step": 9295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7304152725692181,
+      "learning_rate": 1.5492455200096234e-07,
+      "loss": 0.7364,
+      "step": 9296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8101918901000436,
+      "learning_rate": 1.5434742915418466e-07,
+      "loss": 0.7217,
+      "step": 9297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6226630353045093,
+      "learning_rate": 1.5377137489902548e-07,
+      "loss": 0.6549,
+      "step": 9298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8351805596438682,
+      "learning_rate": 1.5319638929800485e-07,
+      "loss": 0.7632,
+      "step": 9299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8081645554575871,
+      "learning_rate": 1.5262247241352945e-07,
+      "loss": 0.6583,
+      "step": 9300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6134327541969915,
+      "learning_rate": 1.520496243078895e-07,
+      "loss": 0.6681,
+      "step": 9301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5380435585216454,
+      "learning_rate": 1.5147784504325746e-07,
+      "loss": 0.6873,
+      "step": 9302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7978338298694088,
+      "learning_rate": 1.509071346816926e-07,
+      "loss": 0.7485,
+      "step": 9303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5582241308265918,
+      "learning_rate": 1.5033749328513537e-07,
+      "loss": 0.7597,
+      "step": 9304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5338763086526532,
+      "learning_rate": 1.4976892091541185e-07,
+      "loss": 0.6083,
+      "step": 9305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8149298068655273,
+      "learning_rate": 1.4920141763423158e-07,
+      "loss": 0.6472,
+      "step": 9306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7713512451374713,
+      "learning_rate": 1.486349835031875e-07,
+      "loss": 0.7545,
+      "step": 9307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6077642775620882,
+      "learning_rate": 1.4806961858375824e-07,
+      "loss": 0.6135,
+      "step": 9308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.695935839933717,
+      "learning_rate": 1.475053229373047e-07,
+      "loss": 0.6744,
+      "step": 9309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7030861479239043,
+      "learning_rate": 1.469420966250723e-07,
+      "loss": 0.7265,
+      "step": 9310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8379397844754426,
+      "learning_rate": 1.4637993970819219e-07,
+      "loss": 0.6841,
+      "step": 9311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.58928195504432,
+      "learning_rate": 1.4581885224767557e-07,
+      "loss": 0.6871,
+      "step": 9312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6598694335921578,
+      "learning_rate": 1.4525883430441922e-07,
+      "loss": 0.7469,
+      "step": 9313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.847375042284769,
+      "learning_rate": 1.446998859392068e-07,
+      "loss": 0.7513,
+      "step": 9314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6196942469957114,
+      "learning_rate": 1.4414200721270198e-07,
+      "loss": 0.7,
+      "step": 9315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.651722372283378,
+      "learning_rate": 1.4358519818545302e-07,
+      "loss": 0.6817,
+      "step": 9316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4620185736229712,
+      "learning_rate": 1.4302945891789487e-07,
+      "loss": 0.6987,
+      "step": 9317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5969563156865756,
+      "learning_rate": 1.4247478947034155e-07,
+      "loss": 0.6432,
+      "step": 9318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7740376885253586,
+      "learning_rate": 1.419211899029971e-07,
+      "loss": 0.7709,
+      "step": 9319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5724361492200893,
+      "learning_rate": 1.4136866027594341e-07,
+      "loss": 0.7302,
+      "step": 9320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.762529061812301,
+      "learning_rate": 1.4081720064915037e-07,
+      "loss": 0.6829,
+      "step": 9321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.732177495747799,
+      "learning_rate": 1.402668110824701e-07,
+      "loss": 0.7528,
+      "step": 9322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5117135004060325,
+      "learning_rate": 1.3971749163563696e-07,
+      "loss": 0.7106,
+      "step": 9323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.936486960015918,
+      "learning_rate": 1.3916924236827444e-07,
+      "loss": 0.7927,
+      "step": 9324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6315407141874998,
+      "learning_rate": 1.3862206333988383e-07,
+      "loss": 0.5763,
+      "step": 9325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6832826870046844,
+      "learning_rate": 1.3807595460985312e-07,
+      "loss": 0.6027,
+      "step": 9326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.633657285099894,
+      "learning_rate": 1.3753091623745497e-07,
+      "loss": 0.7159,
+      "step": 9327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5657399469617896,
+      "learning_rate": 1.3698694828184312e-07,
+      "loss": 0.5689,
+      "step": 9328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5576252990477029,
+      "learning_rate": 1.3644405080205925e-07,
+      "loss": 0.6211,
+      "step": 9329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6803219682719743,
+      "learning_rate": 1.359022238570229e-07,
+      "loss": 0.7426,
+      "step": 9330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7206962427488917,
+      "learning_rate": 1.353614675055437e-07,
+      "loss": 0.7504,
+      "step": 9331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.695239140095105,
+      "learning_rate": 1.3482178180631243e-07,
+      "loss": 0.7181,
+      "step": 9332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5461146588452048,
+      "learning_rate": 1.3428316681790232e-07,
+      "loss": 0.6312,
+      "step": 9333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.634849581054608,
+      "learning_rate": 1.33745622598771e-07,
+      "loss": 0.7263,
+      "step": 9334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7191373351509227,
+      "learning_rate": 1.3320914920726292e-07,
+      "loss": 0.7151,
+      "step": 9335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.806927458293626,
+      "learning_rate": 1.3267374670160037e-07,
+      "loss": 0.7173,
+      "step": 9336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.847358132459803,
+      "learning_rate": 1.3213941513989582e-07,
+      "loss": 0.6338,
+      "step": 9337
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6792896756264375,
+      "learning_rate": 1.316061545801417e-07,
+      "loss": 0.6552,
+      "step": 9338
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7975803650935682,
+      "learning_rate": 1.3107396508021508e-07,
+      "loss": 0.649,
+      "step": 9339
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7320047443084134,
+      "learning_rate": 1.3054284669787754e-07,
+      "loss": 0.757,
+      "step": 9340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7064991477573936,
+      "learning_rate": 1.3001279949077184e-07,
+      "loss": 0.7516,
+      "step": 9341
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8049030315444434,
+      "learning_rate": 1.294838235164275e-07,
+      "loss": 0.7043,
+      "step": 9342
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.783648924170492,
+      "learning_rate": 1.2895591883225754e-07,
+      "loss": 0.7772,
+      "step": 9343
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.661130797786391,
+      "learning_rate": 1.284290854955561e-07,
+      "loss": 0.6693,
+      "step": 9344
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.682829079794363,
+      "learning_rate": 1.27903323563503e-07,
+      "loss": 0.6384,
+      "step": 9345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7792700905516068,
+      "learning_rate": 1.2737863309316257e-07,
+      "loss": 0.6837,
+      "step": 9346
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8187133886049904,
+      "learning_rate": 1.268550141414804e-07,
+      "loss": 0.753,
+      "step": 9347
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.758836973325861,
+      "learning_rate": 1.2633246676528764e-07,
+      "loss": 0.6515,
+      "step": 9348
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7328795167778492,
+      "learning_rate": 1.2581099102129902e-07,
+      "loss": 0.7636,
+      "step": 9349
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8121824781438312,
+      "learning_rate": 1.252905869661114e-07,
+      "loss": 0.6873,
+      "step": 9350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.741948485860588,
+      "learning_rate": 1.2477125465620854e-07,
+      "loss": 0.6576,
+      "step": 9351
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6665539985254938,
+      "learning_rate": 1.242529941479542e-07,
+      "loss": 0.6524,
+      "step": 9352
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6435808873911482,
+      "learning_rate": 1.237358054975979e-07,
+      "loss": 0.7197,
+      "step": 9353
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6404318516170804,
+      "learning_rate": 1.2321968876127244e-07,
+      "loss": 0.723,
+      "step": 9354
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.831279473076429,
+      "learning_rate": 1.2270464399499416e-07,
+      "loss": 0.8039,
+      "step": 9355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6247816181543355,
+      "learning_rate": 1.2219067125466388e-07,
+      "loss": 0.6802,
+      "step": 9356
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7632533356131936,
+      "learning_rate": 1.2167777059606367e-07,
+      "loss": 0.715,
+      "step": 9357
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.656251792043068,
+      "learning_rate": 1.2116594207486232e-07,
+      "loss": 0.5863,
+      "step": 9358
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6262676141905716,
+      "learning_rate": 1.2065518574660983e-07,
+      "loss": 0.653,
+      "step": 9359
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8073241511420874,
+      "learning_rate": 1.201455016667419e-07,
+      "loss": 0.6898,
+      "step": 9360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6183084246280248,
+      "learning_rate": 1.196368898905753e-07,
+      "loss": 0.655,
+      "step": 9361
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.616972179558239,
+      "learning_rate": 1.1912935047331265e-07,
+      "loss": 0.6692,
+      "step": 9362
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7388663060507172,
+      "learning_rate": 1.1862288347004091e-07,
+      "loss": 0.6522,
+      "step": 9363
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5250952476420325,
+      "learning_rate": 1.1811748893572616e-07,
+      "loss": 0.686,
+      "step": 9364
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6662472338063785,
+      "learning_rate": 1.176131669252234e-07,
+      "loss": 0.6984,
+      "step": 9365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5549764713843681,
+      "learning_rate": 1.1710991749326772e-07,
+      "loss": 0.6081,
+      "step": 9366
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5991869130619443,
+      "learning_rate": 1.1660774069447877e-07,
+      "loss": 0.6039,
+      "step": 9367
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7864587149948645,
+      "learning_rate": 1.1610663658336186e-07,
+      "loss": 0.7464,
+      "step": 9368
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9867487428760133,
+      "learning_rate": 1.1560660521430233e-07,
+      "loss": 0.6608,
+      "step": 9369
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6219424473624813,
+      "learning_rate": 1.1510764664157126e-07,
+      "loss": 0.7251,
+      "step": 9370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6957119355691972,
+      "learning_rate": 1.1460976091932307e-07,
+      "loss": 0.6999,
+      "step": 9371
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6944825952036535,
+      "learning_rate": 1.1411294810159457e-07,
+      "loss": 0.6964,
+      "step": 9372
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7535197596326562,
+      "learning_rate": 1.1361720824230704e-07,
+      "loss": 0.6444,
+      "step": 9373
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8297612829240806,
+      "learning_rate": 1.1312254139526635e-07,
+      "loss": 0.7812,
+      "step": 9374
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8623045745422762,
+      "learning_rate": 1.1262894761416066e-07,
+      "loss": 0.797,
+      "step": 9375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.801772300675876,
+      "learning_rate": 1.1213642695256156e-07,
+      "loss": 0.7934,
+      "step": 9376
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6568338935636044,
+      "learning_rate": 1.1164497946392406e-07,
+      "loss": 0.6191,
+      "step": 9377
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.629710990885931,
+      "learning_rate": 1.1115460520158772e-07,
+      "loss": 0.5813,
+      "step": 9378
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6004914877405867,
+      "learning_rate": 1.1066530421877442e-07,
+      "loss": 0.5771,
+      "step": 9379
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7226413707376684,
+      "learning_rate": 1.1017707656859055e-07,
+      "loss": 0.6676,
+      "step": 9380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8615447304953088,
+      "learning_rate": 1.0968992230402598e-07,
+      "loss": 0.7985,
+      "step": 9381
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6921151869689077,
+      "learning_rate": 1.0920384147795172e-07,
+      "loss": 0.7771,
+      "step": 9382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6221705984273835,
+      "learning_rate": 1.0871883414312778e-07,
+      "loss": 0.6666,
+      "step": 9383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5705962467314438,
+      "learning_rate": 1.0823490035218986e-07,
+      "loss": 0.6474,
+      "step": 9384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7104006167842383,
+      "learning_rate": 1.0775204015766483e-07,
+      "loss": 0.8605,
+      "step": 9385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6798442346632074,
+      "learning_rate": 1.0727025361195853e-07,
+      "loss": 0.7488,
+      "step": 9386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8652450432846577,
+      "learning_rate": 1.0678954076736136e-07,
+      "loss": 0.7213,
+      "step": 9387
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7415937054078674,
+      "learning_rate": 1.0630990167604716e-07,
+      "loss": 0.5931,
+      "step": 9388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8481711044022067,
+      "learning_rate": 1.0583133639007203e-07,
+      "loss": 0.706,
+      "step": 9389
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7516193585064288,
+      "learning_rate": 1.053538449613778e-07,
+      "loss": 0.7088,
+      "step": 9390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.620252240786182,
+      "learning_rate": 1.0487742744178964e-07,
+      "loss": 0.6785,
+      "step": 9391
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7365674839200946,
+      "learning_rate": 1.0440208388301399e-07,
+      "loss": 0.732,
+      "step": 9392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.637589589718617,
+      "learning_rate": 1.039278143366429e-07,
+      "loss": 0.6338,
+      "step": 9393
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8376359016842168,
+      "learning_rate": 1.0345461885414965e-07,
+      "loss": 0.6336,
+      "step": 9394
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7308196052245322,
+      "learning_rate": 1.0298249748689204e-07,
+      "loss": 0.7295,
+      "step": 9395
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6808380277400239,
+      "learning_rate": 1.0251145028611464e-07,
+      "loss": 0.6251,
+      "step": 9396
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6294942373354988,
+      "learning_rate": 1.0204147730293767e-07,
+      "loss": 0.6,
+      "step": 9397
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7041566127035381,
+      "learning_rate": 1.0157257858837255e-07,
+      "loss": 0.7238,
+      "step": 9398
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6222572131657285,
+      "learning_rate": 1.0110475419330967e-07,
+      "loss": 0.7622,
+      "step": 9399
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0364432927850973,
+      "learning_rate": 1.0063800416852399e-07,
+      "loss": 0.768,
+      "step": 9400
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7975191300854436,
+      "learning_rate": 1.0017232856467495e-07,
+      "loss": 0.7176,
+      "step": 9401
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5093214742141057,
+      "learning_rate": 9.970772743230329e-08,
+      "loss": 0.649,
+      "step": 9402
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4343705311458053,
+      "learning_rate": 9.924420082183416e-08,
+      "loss": 0.6965,
+      "step": 9403
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6913536743725621,
+      "learning_rate": 9.878174878357738e-08,
+      "loss": 0.6269,
+      "step": 9404
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7215768900124486,
+      "learning_rate": 9.832037136772387e-08,
+      "loss": 0.6705,
+      "step": 9405
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.565162152804664,
+      "learning_rate": 9.786006862434916e-08,
+      "loss": 0.6139,
+      "step": 9406
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6399417139822021,
+      "learning_rate": 9.740084060341104e-08,
+      "loss": 0.6996,
+      "step": 9407
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.671760353975966,
+      "learning_rate": 9.694268735475299e-08,
+      "loss": 0.6743,
+      "step": 9408
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6602976217731225,
+      "learning_rate": 9.648560892809967e-08,
+      "loss": 0.7442,
+      "step": 9409
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7598413490910223,
+      "learning_rate": 9.602960537306027e-08,
+      "loss": 0.7322,
+      "step": 9410
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7576825593121441,
+      "learning_rate": 9.557467673912635e-08,
+      "loss": 0.6904,
+      "step": 9411
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6407909634932698,
+      "learning_rate": 9.512082307567283e-08,
+      "loss": 0.6866,
+      "step": 9412
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8693682646759682,
+      "learning_rate": 9.46680444319581e-08,
+      "loss": 0.7631,
+      "step": 9413
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.78065338467431,
+      "learning_rate": 9.421634085712728e-08,
+      "loss": 0.6334,
+      "step": 9414
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6792038849549156,
+      "learning_rate": 9.376571240020227e-08,
+      "loss": 0.6723,
+      "step": 9415
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.74326436533906,
+      "learning_rate": 9.331615911009284e-08,
+      "loss": 0.682,
+      "step": 9416
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7774904025160487,
+      "learning_rate": 9.286768103559107e-08,
+      "loss": 0.8439,
+      "step": 9417
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6803450951448946,
+      "learning_rate": 9.242027822537247e-08,
+      "loss": 0.6999,
+      "step": 9418
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.626514736365701,
+      "learning_rate": 9.197395072799597e-08,
+      "loss": 0.7507,
+      "step": 9419
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9147094916880236,
+      "learning_rate": 9.152869859190283e-08,
+      "loss": 0.5663,
+      "step": 9420
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.645184270408986,
+      "learning_rate": 9.108452186541771e-08,
+      "loss": 0.7222,
+      "step": 9421
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.745259445087553,
+      "learning_rate": 9.064142059674985e-08,
+      "loss": 0.7458,
+      "step": 9422
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6908506915448078,
+      "learning_rate": 9.019939483399076e-08,
+      "loss": 0.6719,
+      "step": 9423
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8034257047030355,
+      "learning_rate": 8.975844462511652e-08,
+      "loss": 0.7185,
+      "step": 9424
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7462225913889697,
+      "learning_rate": 8.931857001798216e-08,
+      "loss": 0.6562,
+      "step": 9425
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.700078979326214,
+      "learning_rate": 8.887977106033285e-08,
+      "loss": 0.6466,
+      "step": 9426
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8522994307817322,
+      "learning_rate": 8.84420477997916e-08,
+      "loss": 0.7504,
+      "step": 9427
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6080686398108495,
+      "learning_rate": 8.800540028386595e-08,
+      "loss": 0.6406,
+      "step": 9428
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7004782602843376,
+      "learning_rate": 8.756982855994911e-08,
+      "loss": 0.6996,
+      "step": 9429
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8207728151070466,
+      "learning_rate": 8.713533267531326e-08,
+      "loss": 0.7679,
+      "step": 9430
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7478520752621098,
+      "learning_rate": 8.670191267711736e-08,
+      "loss": 0.715,
+      "step": 9431
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.66706241086095,
+      "learning_rate": 8.626956861240265e-08,
+      "loss": 0.6582,
+      "step": 9432
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6439981567690312,
+      "learning_rate": 8.58383005280916e-08,
+      "loss": 0.5828,
+      "step": 9433
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7405589455947643,
+      "learning_rate": 8.540810847099345e-08,
+      "loss": 0.7158,
+      "step": 9434
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7367774924168071,
+      "learning_rate": 8.497899248779862e-08,
+      "loss": 0.757,
+      "step": 9435
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6737570175181034,
+      "learning_rate": 8.455095262508095e-08,
+      "loss": 0.7011,
+      "step": 9436
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6452133588641074,
+      "learning_rate": 8.412398892929663e-08,
+      "loss": 0.6645,
+      "step": 9437
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.681366338145333,
+      "learning_rate": 8.369810144678636e-08,
+      "loss": 0.77,
+      "step": 9438
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.550502000121852,
+      "learning_rate": 8.327329022377317e-08,
+      "loss": 0.6442,
+      "step": 9439
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6022409815852374,
+      "learning_rate": 8.284955530636462e-08,
+      "loss": 0.6408,
+      "step": 9440
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5452686011312082,
+      "learning_rate": 8.242689674054949e-08,
+      "loss": 0.6169,
+      "step": 9441
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8167821762833238,
+      "learning_rate": 8.200531457220218e-08,
+      "loss": 0.7093,
+      "step": 9442
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7396942787558083,
+      "learning_rate": 8.15848088470772e-08,
+      "loss": 0.7135,
+      "step": 9443
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9776231560210884,
+      "learning_rate": 8.116537961081473e-08,
+      "loss": 0.6921,
+      "step": 9444
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.75401788492991,
+      "learning_rate": 8.074702690893722e-08,
+      "loss": 0.6245,
+      "step": 9445
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.638899199915196,
+      "learning_rate": 8.032975078684945e-08,
+      "loss": 0.6686,
+      "step": 9446
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6823860072883687,
+      "learning_rate": 7.99135512898408e-08,
+      "loss": 0.7626,
+      "step": 9447
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5854351031807767,
+      "learning_rate": 7.949842846308398e-08,
+      "loss": 0.6313,
+      "step": 9448
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5310605391572794,
+      "learning_rate": 7.908438235163407e-08,
+      "loss": 0.5886,
+      "step": 9449
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8708623272585712,
+      "learning_rate": 7.867141300042736e-08,
+      "loss": 0.7019,
+      "step": 9450
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3911077288709224,
+      "learning_rate": 7.825952045428797e-08,
+      "loss": 0.5764,
+      "step": 9451
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6768022829466143,
+      "learning_rate": 7.784870475791794e-08,
+      "loss": 0.6353,
+      "step": 9452
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5242529784224401,
+      "learning_rate": 7.743896595590605e-08,
+      "loss": 0.6805,
+      "step": 9453
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8249475019113455,
+      "learning_rate": 7.703030409272339e-08,
+      "loss": 0.744,
+      "step": 9454
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7586110111232993,
+      "learning_rate": 7.662271921272224e-08,
+      "loss": 0.6442,
+      "step": 9455
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7143857110798468,
+      "learning_rate": 7.621621136014168e-08,
+      "loss": 0.5942,
+      "step": 9456
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6898773689992317,
+      "learning_rate": 7.581078057909974e-08,
+      "loss": 0.6309,
+      "step": 9457
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.634801653449755,
+      "learning_rate": 7.540642691360123e-08,
+      "loss": 0.6671,
+      "step": 9458
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8680966984373397,
+      "learning_rate": 7.500315040753214e-08,
+      "loss": 0.6873,
+      "step": 9459
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6719597635551595,
+      "learning_rate": 7.460095110466192e-08,
+      "loss": 0.6266,
+      "step": 9460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6156246547794542,
+      "learning_rate": 7.41998290486412e-08,
+      "loss": 0.6576,
+      "step": 9461
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4988781002438796,
+      "learning_rate": 7.379978428300738e-08,
+      "loss": 0.5995,
+      "step": 9462
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6309987376080932,
+      "learning_rate": 7.340081685117906e-08,
+      "loss": 0.6821,
+      "step": 9463
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8204765391454656,
+      "learning_rate": 7.300292679645716e-08,
+      "loss": 0.715,
+      "step": 9464
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.781509439351711,
+      "learning_rate": 7.260611416202712e-08,
+      "loss": 0.8175,
+      "step": 9465
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7769442386139096,
+      "learning_rate": 7.221037899095561e-08,
+      "loss": 0.6977,
+      "step": 9466
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.884030339551365,
+      "learning_rate": 7.181572132619385e-08,
+      "loss": 0.645,
+      "step": 9467
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5657382868794574,
+      "learning_rate": 7.142214121057755e-08,
+      "loss": 0.6959,
+      "step": 9468
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7086982932887396,
+      "learning_rate": 7.102963868682034e-08,
+      "loss": 0.5867,
+      "step": 9469
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4962641401982082,
+      "learning_rate": 7.063821379752589e-08,
+      "loss": 0.5343,
+      "step": 9470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7827013511340437,
+      "learning_rate": 7.024786658517468e-08,
+      "loss": 0.7294,
+      "step": 9471
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7029572387275005,
+      "learning_rate": 6.985859709213283e-08,
+      "loss": 0.6746,
+      "step": 9472
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.528253388269664,
+      "learning_rate": 6.947040536065208e-08,
+      "loss": 0.6944,
+      "step": 9473
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.866562625793677,
+      "learning_rate": 6.908329143286096e-08,
+      "loss": 0.6758,
+      "step": 9474
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6755032350387424,
+      "learning_rate": 6.869725535077698e-08,
+      "loss": 0.7056,
+      "step": 9475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7746080634624017,
+      "learning_rate": 6.831229715629884e-08,
+      "loss": 0.6918,
+      "step": 9476
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.710479335270481,
+      "learning_rate": 6.792841689120533e-08,
+      "loss": 0.8261,
+      "step": 9477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.633912267250495,
+      "learning_rate": 6.754561459716202e-08,
+      "loss": 0.7372,
+      "step": 9478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8059470626753045,
+      "learning_rate": 6.716389031571568e-08,
+      "loss": 0.6856,
+      "step": 9479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7177100237104101,
+      "learning_rate": 6.678324408829762e-08,
+      "loss": 0.6838,
+      "step": 9480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6373046738415207,
+      "learning_rate": 6.640367595622033e-08,
+      "loss": 0.7585,
+      "step": 9481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8719583939143258,
+      "learning_rate": 6.602518596067975e-08,
+      "loss": 0.6701,
+      "step": 9482
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7507052228301732,
+      "learning_rate": 6.564777414275525e-08,
+      "loss": 0.7428,
+      "step": 9483
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6062017308145797,
+      "learning_rate": 6.52714405434085e-08,
+      "loss": 0.65,
+      "step": 9484
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8043004451790021,
+      "learning_rate": 6.489618520348573e-08,
+      "loss": 0.7727,
+      "step": 9485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6345253892914255,
+      "learning_rate": 6.452200816371435e-08,
+      "loss": 0.7133,
+      "step": 9486
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6192129197947587,
+      "learning_rate": 6.41489094647052e-08,
+      "loss": 0.705,
+      "step": 9487
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7888330879688503,
+      "learning_rate": 6.377688914695256e-08,
+      "loss": 0.7332,
+      "step": 9488
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5641196925584768,
+      "learning_rate": 6.340594725083415e-08,
+      "loss": 0.734,
+      "step": 9489
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9912579431641058,
+      "learning_rate": 6.303608381660887e-08,
+      "loss": 0.7217,
+      "step": 9490
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5776840514556638,
+      "learning_rate": 6.266729888442013e-08,
+      "loss": 0.6663,
+      "step": 9491
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7249885914365968,
+      "learning_rate": 6.229959249429263e-08,
+      "loss": 0.6978,
+      "step": 9492
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8158158270080882,
+      "learning_rate": 6.193296468613663e-08,
+      "loss": 0.6448,
+      "step": 9493
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6464321238990238,
+      "learning_rate": 6.156741549974365e-08,
+      "loss": 0.6866,
+      "step": 9494
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6656953848178833,
+      "learning_rate": 6.120294497478752e-08,
+      "loss": 0.6183,
+      "step": 9495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9088722193655174,
+      "learning_rate": 6.083955315082657e-08,
+      "loss": 0.6723,
+      "step": 9496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5716830550858858,
+      "learning_rate": 6.04772400673015e-08,
+      "loss": 0.6995,
+      "step": 9497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.792143791691787,
+      "learning_rate": 6.011600576353416e-08,
+      "loss": 0.7165,
+      "step": 9498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6182212968308853,
+      "learning_rate": 5.97558502787332e-08,
+      "loss": 0.7434,
+      "step": 9499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7037892359301718,
+      "learning_rate": 5.939677365198626e-08,
+      "loss": 0.6117,
+      "step": 9500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.707700336540344,
+      "learning_rate": 5.903877592226548e-08,
+      "loss": 0.7474,
+      "step": 9501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6940925472575223,
+      "learning_rate": 5.868185712842645e-08,
+      "loss": 0.6836,
+      "step": 9502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6486427451514054,
+      "learning_rate": 5.832601730920706e-08,
+      "loss": 0.7138,
+      "step": 9503
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5679311999282515,
+      "learning_rate": 5.797125650322866e-08,
+      "loss": 0.5144,
+      "step": 9504
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6840151836203492,
+      "learning_rate": 5.7617574748993764e-08,
+      "loss": 0.6978,
+      "step": 9505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7141209646309552,
+      "learning_rate": 5.726497208488946e-08,
+      "loss": 0.6544,
+      "step": 9506
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7961651134883347,
+      "learning_rate": 5.691344854918623e-08,
+      "loss": 0.8223,
+      "step": 9507
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.721568167980704,
+      "learning_rate": 5.6563004180034685e-08,
+      "loss": 0.7074,
+      "step": 9508
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7756903237314543,
+      "learning_rate": 5.6213639015472166e-08,
+      "loss": 0.729,
+      "step": 9509
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7530493444827209,
+      "learning_rate": 5.586535309341501e-08,
+      "loss": 0.6859,
+      "step": 9510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8363965552736352,
+      "learning_rate": 5.55181464516652e-08,
+      "loss": 0.7653,
+      "step": 9511
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.632990922923886,
+      "learning_rate": 5.517201912790593e-08,
+      "loss": 0.6096,
+      "step": 9512
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6297590626113077,
+      "learning_rate": 5.4826971159704925e-08,
+      "loss": 0.6823,
+      "step": 9513
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.001130699233612,
+      "learning_rate": 5.448300258451111e-08,
+      "loss": 0.7525,
+      "step": 9514
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5364424660076623,
+      "learning_rate": 5.4140113439655753e-08,
+      "loss": 0.6215,
+      "step": 9515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7204574771548173,
+      "learning_rate": 5.379830376235573e-08,
+      "loss": 0.6571,
+      "step": 9516
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6017778895293966,
+      "learning_rate": 5.3457573589709156e-08,
+      "loss": 0.682,
+      "step": 9517
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5577355583757446,
+      "learning_rate": 5.311792295869644e-08,
+      "loss": 0.63,
+      "step": 9518
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.682049891016354,
+      "learning_rate": 5.2779351906181445e-08,
+      "loss": 0.6591,
+      "step": 9519
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5620593916862504,
+      "learning_rate": 5.24418604689092e-08,
+      "loss": 0.5631,
+      "step": 9520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.681310877232562,
+      "learning_rate": 5.210544868351153e-08,
+      "loss": 0.5983,
+      "step": 9521
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8030213916103963,
+      "learning_rate": 5.177011658650033e-08,
+      "loss": 0.6789,
+      "step": 9522
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8096040261391229,
+      "learning_rate": 5.143586421426982e-08,
+      "loss": 0.6906,
+      "step": 9523
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.836173510001181,
+      "learning_rate": 5.1102691603097664e-08,
+      "loss": 0.7915,
+      "step": 9524
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6777462206542368,
+      "learning_rate": 5.077059878914492e-08,
+      "loss": 0.7249,
+      "step": 9525
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7752150890688336,
+      "learning_rate": 5.043958580845498e-08,
+      "loss": 0.6774,
+      "step": 9526
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.935721956313621,
+      "learning_rate": 5.010965269695578e-08,
+      "loss": 0.7067,
+      "step": 9527
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6102260340258132,
+      "learning_rate": 4.978079949045311e-08,
+      "loss": 0.7253,
+      "step": 9528
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.659590054401639,
+      "learning_rate": 4.945302622464177e-08,
+      "loss": 0.619,
+      "step": 9529
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5797987844636674,
+      "learning_rate": 4.912633293509439e-08,
+      "loss": 0.7114,
+      "step": 9530
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.651232052055731,
+      "learning_rate": 4.8800719657270404e-08,
+      "loss": 0.6429,
+      "step": 9531
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7290046370594587,
+      "learning_rate": 4.84761864265082e-08,
+      "loss": 0.734,
+      "step": 9532
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7180351129100842,
+      "learning_rate": 4.815273327803183e-08,
+      "loss": 0.6383,
+      "step": 9533
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5391683968772378,
+      "learning_rate": 4.783036024694543e-08,
+      "loss": 0.6444,
+      "step": 9534
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7809726189745998,
+      "learning_rate": 4.750906736824101e-08,
+      "loss": 0.6574,
+      "step": 9535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6544166039332864,
+      "learning_rate": 4.7188854676786246e-08,
+      "loss": 0.7301,
+      "step": 9536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5079155368243176,
+      "learning_rate": 4.6869722207337763e-08,
+      "loss": 0.6591,
+      "step": 9537
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6571193553186434,
+      "learning_rate": 4.6551669994531204e-08,
+      "loss": 0.6755,
+      "step": 9538
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7825879578730972,
+      "learning_rate": 4.623469807288561e-08,
+      "loss": 0.7396,
+      "step": 9539
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7840446572194073,
+      "learning_rate": 4.591880647680458e-08,
+      "loss": 0.657,
+      "step": 9540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9065872718206953,
+      "learning_rate": 4.5603995240572906e-08,
+      "loss": 0.7672,
+      "step": 9541
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.696836293573375,
+      "learning_rate": 4.529026439835771e-08,
+      "loss": 0.72,
+      "step": 9542
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.739553332302951,
+      "learning_rate": 4.4977613984210634e-08,
+      "loss": 0.6855,
+      "step": 9543
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6854862459503848,
+      "learning_rate": 4.4666044032063425e-08,
+      "loss": 0.7009,
+      "step": 9544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6963922444075805,
+      "learning_rate": 4.4355554575734594e-08,
+      "loss": 0.6945,
+      "step": 9545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5800756558416964,
+      "learning_rate": 4.404614564892051e-08,
+      "loss": 0.588,
+      "step": 9546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6478454303775683,
+      "learning_rate": 4.373781728520321e-08,
+      "loss": 0.6437,
+      "step": 9547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5023842441634008,
+      "learning_rate": 4.3430569518048135e-08,
+      "loss": 0.6218,
+      "step": 9548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5924937602995637,
+      "learning_rate": 4.3124402380800846e-08,
+      "loss": 0.6583,
+      "step": 9549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6325314575288126,
+      "learning_rate": 4.281931590669253e-08,
+      "loss": 0.5673,
+      "step": 9550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5829462492317814,
+      "learning_rate": 4.251531012883337e-08,
+      "loss": 0.6106,
+      "step": 9551
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5521949626144838,
+      "learning_rate": 4.2212385080220295e-08,
+      "loss": 0.6161,
+      "step": 9552
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5820619083433027,
+      "learning_rate": 4.191054079373036e-08,
+      "loss": 0.7234,
+      "step": 9553
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6808104992758746,
+      "learning_rate": 4.160977730212401e-08,
+      "loss": 0.6616,
+      "step": 9554
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.924873928536549,
+      "learning_rate": 4.131009463804403e-08,
+      "loss": 0.8662,
+      "step": 9555
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4615127567216981,
+      "learning_rate": 4.101149283401773e-08,
+      "loss": 0.7099,
+      "step": 9556
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.615602544070035,
+      "learning_rate": 4.071397192245252e-08,
+      "loss": 0.6222,
+      "step": 9557
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.954739249642568,
+      "learning_rate": 4.041753193563924e-08,
+      "loss": 0.7783,
+      "step": 9558
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6038174186370082,
+      "learning_rate": 4.0122172905753264e-08,
+      "loss": 0.7212,
+      "step": 9559
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5473590138448612,
+      "learning_rate": 3.982789486485006e-08,
+      "loss": 0.5879,
+      "step": 9560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5495334233350668,
+      "learning_rate": 3.953469784486852e-08,
+      "loss": 0.6746,
+      "step": 9561
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5247979141054615,
+      "learning_rate": 3.924258187763208e-08,
+      "loss": 0.6481,
+      "step": 9562
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.923834001273207,
+      "learning_rate": 3.895154699484427e-08,
+      "loss": 0.7229,
+      "step": 9563
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9727940095964407,
+      "learning_rate": 3.866159322809315e-08,
+      "loss": 0.7823,
+      "step": 9564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6465939082770047,
+      "learning_rate": 3.8372720608848e-08,
+      "loss": 0.7266,
+      "step": 9565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6971371645450848,
+      "learning_rate": 3.808492916846041e-08,
+      "loss": 0.7288,
+      "step": 9566
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7647054382027547,
+      "learning_rate": 3.77982189381676e-08,
+      "loss": 0.6905,
+      "step": 9567
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7337785020115393,
+      "learning_rate": 3.751258994908691e-08,
+      "loss": 0.7073,
+      "step": 9568
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9913478597915608,
+      "learning_rate": 3.72280422322191e-08,
+      "loss": 0.7697,
+      "step": 9569
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7756431875942662,
+      "learning_rate": 3.6944575818446126e-08,
+      "loss": 0.69,
+      "step": 9570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6794395528746324,
+      "learning_rate": 3.6662190738535606e-08,
+      "loss": 0.7857,
+      "step": 9571
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7678436916808111,
+      "learning_rate": 3.638088702313414e-08,
+      "loss": 0.646,
+      "step": 9572
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7993167160166388,
+      "learning_rate": 3.610066470277507e-08,
+      "loss": 0.676,
+      "step": 9573
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5841152046243077,
+      "learning_rate": 3.582152380786963e-08,
+      "loss": 0.5893,
+      "step": 9574
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6703393696365203,
+      "learning_rate": 3.554346436871581e-08,
+      "loss": 0.7292,
+      "step": 9575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6449459827254167,
+      "learning_rate": 3.52664864154928e-08,
+      "loss": 0.6117,
+      "step": 9576
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9219088939810345,
+      "learning_rate": 3.499058997826099e-08,
+      "loss": 0.7298,
+      "step": 9577
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5935011107809127,
+      "learning_rate": 3.471577508696533e-08,
+      "loss": 0.5968,
+      "step": 9578
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5925932403139895,
+      "learning_rate": 3.444204177143307e-08,
+      "loss": 0.623,
+      "step": 9579
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.782416756084071,
+      "learning_rate": 3.416939006137265e-08,
+      "loss": 0.7689,
+      "step": 9580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7447214103114526,
+      "learning_rate": 3.389781998637709e-08,
+      "loss": 0.6998,
+      "step": 9581
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6536370107942204,
+      "learning_rate": 3.3627331575921686e-08,
+      "loss": 0.6672,
+      "step": 9582
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7187325303429446,
+      "learning_rate": 3.3357924859361845e-08,
+      "loss": 0.7012,
+      "step": 9583
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8955684515759144,
+      "learning_rate": 3.3089599865938625e-08,
+      "loss": 0.6829,
+      "step": 9584
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7153222583829413,
+      "learning_rate": 3.282235662477429e-08,
+      "loss": 0.7687,
+      "step": 9585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.770308421957577,
+      "learning_rate": 3.2556195164873405e-08,
+      "loss": 0.6477,
+      "step": 9586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.629486287642064,
+      "learning_rate": 3.2291115515125093e-08,
+      "loss": 0.6571,
+      "step": 9587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7828578010716696,
+      "learning_rate": 3.202711770429745e-08,
+      "loss": 0.7029,
+      "step": 9588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.783499374620443,
+      "learning_rate": 3.176420176104533e-08,
+      "loss": 0.6968,
+      "step": 9589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7423644115712555,
+      "learning_rate": 3.1502367713903695e-08,
+      "loss": 0.7301,
+      "step": 9590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7655218329922557,
+      "learning_rate": 3.12416155912898e-08,
+      "loss": 0.6931,
+      "step": 9591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7484148997191309,
+      "learning_rate": 3.0981945421504345e-08,
+      "loss": 0.742,
+      "step": 9592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6889592247363676,
+      "learning_rate": 3.0723357232731455e-08,
+      "loss": 0.6322,
+      "step": 9593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6742068227188056,
+      "learning_rate": 3.0465851053035345e-08,
+      "loss": 0.5973,
+      "step": 9594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.658300994042665,
+      "learning_rate": 3.0209426910364766e-08,
+      "loss": 0.7638,
+      "step": 9595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.561736898588684,
+      "learning_rate": 2.995408483255191e-08,
+      "loss": 0.6049,
+      "step": 9596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7656308985229812,
+      "learning_rate": 2.9699824847307933e-08,
+      "loss": 0.7248,
+      "step": 9597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5919797364329935,
+      "learning_rate": 2.9446646982230763e-08,
+      "loss": 0.6782,
+      "step": 9598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7407948021472628,
+      "learning_rate": 2.919455126479731e-08,
+      "loss": 0.7177,
+      "step": 9599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6195766093351134,
+      "learning_rate": 2.894353772237013e-08,
+      "loss": 0.567,
+      "step": 9600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8401774040526315,
+      "learning_rate": 2.8693606382191876e-08,
+      "loss": 0.8156,
+      "step": 9601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.586921656800293,
+      "learning_rate": 2.8444757271388625e-08,
+      "loss": 0.7336,
+      "step": 9602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7564114273771738,
+      "learning_rate": 2.819699041696877e-08,
+      "loss": 0.7562,
+      "step": 9603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7466936053595925,
+      "learning_rate": 2.7950305845825254e-08,
+      "loss": 0.6833,
+      "step": 9604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7033376017472277,
+      "learning_rate": 2.7704703584729985e-08,
+      "loss": 0.7689,
+      "step": 9605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5863377258152949,
+      "learning_rate": 2.7460183660339422e-08,
+      "loss": 0.6984,
+      "step": 9606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0272524378790626,
+      "learning_rate": 2.7216746099193448e-08,
+      "loss": 0.6807,
+      "step": 9607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5975543790019915,
+      "learning_rate": 2.6974390927712034e-08,
+      "loss": 0.6843,
+      "step": 9608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6388291697059447,
+      "learning_rate": 2.6733118172200812e-08,
+      "loss": 0.6857,
+      "step": 9609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5592279432842686,
+      "learning_rate": 2.6492927858844386e-08,
+      "loss": 0.6555,
+      "step": 9610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6034240035241638,
+      "learning_rate": 2.6253820013713017e-08,
+      "loss": 0.6381,
+      "step": 9611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6946540536432775,
+      "learning_rate": 2.6015794662757055e-08,
+      "loss": 0.6846,
+      "step": 9612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7225744072643763,
+      "learning_rate": 2.5778851831811392e-08,
+      "loss": 0.7251,
+      "step": 9613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6038628126765848,
+      "learning_rate": 2.554299154659212e-08,
+      "loss": 0.6683,
+      "step": 9614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.673729472783224,
+      "learning_rate": 2.5308213832697658e-08,
+      "loss": 0.7261,
+      "step": 9615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6990401901929215,
+      "learning_rate": 2.5074518715609843e-08,
+      "loss": 0.7117,
+      "step": 9616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6665138770587684,
+      "learning_rate": 2.4841906220692825e-08,
+      "loss": 0.6896,
+      "step": 9617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7577331276681882,
+      "learning_rate": 2.461037637319308e-08,
+      "loss": 0.6992,
+      "step": 9618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.832379629273432,
+      "learning_rate": 2.4379929198238285e-08,
+      "loss": 0.7282,
+      "step": 9619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.752051759087949,
+      "learning_rate": 2.4150564720841763e-08,
+      "loss": 0.748,
+      "step": 9620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7443054920924745,
+      "learning_rate": 2.3922282965896936e-08,
+      "loss": 0.6948,
+      "step": 9621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.563872679925648,
+      "learning_rate": 2.3695083958179543e-08,
+      "loss": 0.654,
+      "step": 9622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.789764500311137,
+      "learning_rate": 2.3468967722347635e-08,
+      "loss": 0.7633,
+      "step": 9623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7614250422498046,
+      "learning_rate": 2.3243934282944912e-08,
+      "loss": 0.6923,
+      "step": 9624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7086680323811136,
+      "learning_rate": 2.3019983664394064e-08,
+      "loss": 0.7576,
+      "step": 9625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8243468464401964,
+      "learning_rate": 2.2797115891002308e-08,
+      "loss": 0.7506,
+      "step": 9626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8382359181164998,
+      "learning_rate": 2.2575330986956968e-08,
+      "loss": 0.6899,
+      "step": 9627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.711964555016779,
+      "learning_rate": 2.2354628976328786e-08,
+      "loss": 0.6424,
+      "step": 9628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6574125129336976,
+      "learning_rate": 2.2135009883074155e-08,
+      "loss": 0.5457,
+      "step": 9629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.470929810445323,
+      "learning_rate": 2.1916473731027343e-08,
+      "loss": 0.6801,
+      "step": 9630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7670042059843676,
+      "learning_rate": 2.1699020543907157e-08,
+      "loss": 0.6014,
+      "step": 9631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8821606786857161,
+      "learning_rate": 2.1482650345315826e-08,
+      "loss": 0.665,
+      "step": 9632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6664184114715632,
+      "learning_rate": 2.1267363158735676e-08,
+      "loss": 0.7083,
+      "step": 9633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.705911715860228,
+      "learning_rate": 2.1053159007533575e-08,
+      "loss": 0.6824,
+      "step": 9634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7456083344015636,
+      "learning_rate": 2.0840037914958698e-08,
+      "loss": 0.6966,
+      "step": 9635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6545046167332191,
+      "learning_rate": 2.0627999904139218e-08,
+      "loss": 0.6038,
+      "step": 9636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8839922766007815,
+      "learning_rate": 2.041704499809227e-08,
+      "loss": 0.7633,
+      "step": 9637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6442871165201445,
+      "learning_rate": 2.0207173219710664e-08,
+      "loss": 0.7207,
+      "step": 9638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8057249181482182,
+      "learning_rate": 1.9998384591773945e-08,
+      "loss": 0.7916,
+      "step": 9639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6379867021567172,
+      "learning_rate": 1.979067913694399e-08,
+      "loss": 0.714,
+      "step": 9640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7978173995445093,
+      "learning_rate": 1.9584056877761658e-08,
+      "loss": 0.6308,
+      "step": 9641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8522532972135437,
+      "learning_rate": 1.9378517836653454e-08,
+      "loss": 0.6179,
+      "step": 9642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6906803348048336,
+      "learning_rate": 1.91740620359282e-08,
+      "loss": 0.6019,
+      "step": 9643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7531966996092696,
+      "learning_rate": 1.8970689497775917e-08,
+      "loss": 0.675,
+      "step": 9644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.728651057249247,
+      "learning_rate": 1.876840024427007e-08,
+      "loss": 0.7092,
+      "step": 9645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5379281297714842,
+      "learning_rate": 1.85671942973642e-08,
+      "loss": 0.6308,
+      "step": 9646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7675827172247032,
+      "learning_rate": 1.8367071678897507e-08,
+      "loss": 0.6974,
+      "step": 9647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7158119619808374,
+      "learning_rate": 1.8168032410590398e-08,
+      "loss": 0.6727,
+      "step": 9648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.643652158228756,
+      "learning_rate": 1.7970076514044476e-08,
+      "loss": 0.8226,
+      "step": 9649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.877227745255658,
+      "learning_rate": 1.7773204010745892e-08,
+      "loss": 0.6955,
+      "step": 9650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8893676527357877,
+      "learning_rate": 1.7577414922060888e-08,
+      "loss": 0.7242,
+      "step": 9651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.627296937520523,
+      "learning_rate": 1.738270926924024e-08,
+      "loss": 0.6402,
+      "step": 9652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.616663575398968,
+      "learning_rate": 1.7189087073415933e-08,
+      "loss": 0.7571,
+      "step": 9653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6853571802209921,
+      "learning_rate": 1.699654835560116e-08,
+      "loss": 0.6763,
+      "step": 9654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.854656056250111,
+      "learning_rate": 1.680509313669587e-08,
+      "loss": 0.6867,
+      "step": 9655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4804921632252286,
+      "learning_rate": 1.6614721437477887e-08,
+      "loss": 0.6332,
+      "step": 9656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6700954955835952,
+      "learning_rate": 1.642543327860846e-08,
+      "loss": 0.6996,
+      "step": 9657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.592121528151136,
+      "learning_rate": 1.6237228680633376e-08,
+      "loss": 0.6434,
+      "step": 9658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6533858632363638,
+      "learning_rate": 1.605010766397741e-08,
+      "loss": 0.8153,
+      "step": 9659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.804843545317245,
+      "learning_rate": 1.586407024895209e-08,
+      "loss": 0.6773,
+      "step": 9660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4865888180732818,
+      "learning_rate": 1.5679116455746823e-08,
+      "loss": 0.6879,
+      "step": 9661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7862992328554177,
+      "learning_rate": 1.5495246304435552e-08,
+      "loss": 0.7552,
+      "step": 9662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8059227397518534,
+      "learning_rate": 1.5312459814975644e-08,
+      "loss": 0.6882,
+      "step": 9663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5367676202214664,
+      "learning_rate": 1.5130757007205676e-08,
+      "loss": 0.5667,
+      "step": 9664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.836586511875315,
+      "learning_rate": 1.495013790084654e-08,
+      "loss": 0.7366,
+      "step": 9665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7310378350968094,
+      "learning_rate": 1.4770602515500332e-08,
+      "loss": 0.6892,
+      "step": 9666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7893982283088452,
+      "learning_rate": 1.4592150870653688e-08,
+      "loss": 0.7379,
+      "step": 9667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6391847665314565,
+      "learning_rate": 1.4414782985674447e-08,
+      "loss": 0.6567,
+      "step": 9668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.648479239098497,
+      "learning_rate": 1.4238498879813878e-08,
+      "loss": 0.6808,
+      "step": 9669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6315940355015648,
+      "learning_rate": 1.4063298572204454e-08,
+      "loss": 0.7107,
+      "step": 9670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.652497062929891,
+      "learning_rate": 1.3889182081860963e-08,
+      "loss": 0.7429,
+      "step": 9671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.646812479137008,
+      "learning_rate": 1.3716149427682734e-08,
+      "loss": 0.6522,
+      "step": 9672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6886811009695795,
+      "learning_rate": 1.3544200628446968e-08,
+      "loss": 0.6619,
+      "step": 9673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7001777649396592,
+      "learning_rate": 1.3373335702818735e-08,
+      "loss": 0.7377,
+      "step": 9674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5931986521862718,
+      "learning_rate": 1.3203554669339868e-08,
+      "loss": 0.6784,
+      "step": 9675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5438624429645642,
+      "learning_rate": 1.3034857546441182e-08,
+      "loss": 0.5341,
+      "step": 9676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4887640945970237,
+      "learning_rate": 1.2867244352428033e-08,
+      "loss": 0.6961,
+      "step": 9677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7453192466708156,
+      "learning_rate": 1.2700715105495865e-08,
+      "loss": 0.6643,
+      "step": 9678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9154223364917178,
+      "learning_rate": 1.2535269823716889e-08,
+      "loss": 0.7022,
+      "step": 9679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6207314273673628,
+      "learning_rate": 1.2370908525046744e-08,
+      "loss": 0.7324,
+      "step": 9680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.781701331962845,
+      "learning_rate": 1.2207631227326711e-08,
+      "loss": 0.6468,
+      "step": 9681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.766644934566483,
+      "learning_rate": 1.2045437948275952e-08,
+      "loss": 0.7227,
+      "step": 9682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7699821522867707,
+      "learning_rate": 1.188432870549927e-08,
+      "loss": 0.7304,
+      "step": 9683
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7396705355516582,
+      "learning_rate": 1.1724303516481572e-08,
+      "loss": 0.6502,
+      "step": 9684
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6905569226968575,
+      "learning_rate": 1.1565362398592295e-08,
+      "loss": 0.8001,
+      "step": 9685
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6747194976240847,
+      "learning_rate": 1.1407505369080973e-08,
+      "loss": 0.6568,
+      "step": 9686
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.642429861783825,
+      "learning_rate": 1.1250732445080569e-08,
+      "loss": 0.6395,
+      "step": 9687
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6875176860247043,
+      "learning_rate": 1.1095043643606363e-08,
+      "loss": 0.7886,
+      "step": 9688
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8228693514058845,
+      "learning_rate": 1.0940438981555945e-08,
+      "loss": 0.7219,
+      "step": 9689
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7675485509518638,
+      "learning_rate": 1.0786918475710339e-08,
+      "loss": 0.6434,
+      "step": 9690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.647818234835014,
+      "learning_rate": 1.0634482142730662e-08,
+      "loss": 0.7144,
+      "step": 9691
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7718676572281018,
+      "learning_rate": 1.0483129999161456e-08,
+      "loss": 0.6408,
+      "step": 9692
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6318067602617774,
+      "learning_rate": 1.0332862061429583e-08,
+      "loss": 0.6004,
+      "step": 9693
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6351165397625582,
+      "learning_rate": 1.0183678345845328e-08,
+      "loss": 0.7552,
+      "step": 9694
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5644902946973864,
+      "learning_rate": 1.0035578868600182e-08,
+      "loss": 0.6444,
+      "step": 9695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6428946667146398,
+      "learning_rate": 9.888563645765736e-09,
+      "loss": 0.5811,
+      "step": 9696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5742527303659848,
+      "learning_rate": 9.742632693301445e-09,
+      "loss": 0.692,
+      "step": 9697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7736699263772455,
+      "learning_rate": 9.597786027042422e-09,
+      "loss": 0.7327,
+      "step": 9698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9491332009997895,
+      "learning_rate": 9.454023662712752e-09,
+      "loss": 0.6928,
+      "step": 9699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7185475542191888,
+      "learning_rate": 9.311345615913291e-09,
+      "loss": 0.6323,
+      "step": 9700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5105438695817894,
+      "learning_rate": 9.169751902131652e-09,
+      "loss": 0.6499,
+      "step": 9701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.514827773029279,
+      "learning_rate": 9.029242536733318e-09,
+      "loss": 0.6234,
+      "step": 9702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6461197241040941,
+      "learning_rate": 8.889817534969425e-09,
+      "loss": 0.6747,
+      "step": 9703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7957496588047202,
+      "learning_rate": 8.751476911972313e-09,
+      "loss": 0.7837,
+      "step": 9704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6629104613611898,
+      "learning_rate": 8.614220682756635e-09,
+      "loss": 0.6097,
+      "step": 9705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6919638021174255,
+      "learning_rate": 8.478048862219368e-09,
+      "loss": 0.7037,
+      "step": 9706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5271956567781249,
+      "learning_rate": 8.342961465140908e-09,
+      "loss": 0.5894,
+      "step": 9707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6611920184684048,
+      "learning_rate": 8.208958506181752e-09,
+      "loss": 0.7251,
+      "step": 9708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7030516189999174,
+      "learning_rate": 8.076039999885821e-09,
+      "loss": 0.7094,
+      "step": 9709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5774737119575863,
+      "learning_rate": 7.944205960678242e-09,
+      "loss": 0.6555,
+      "step": 9710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5424960825862553,
+      "learning_rate": 7.813456402870901e-09,
+      "loss": 0.7548,
+      "step": 9711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9115060508268176,
+      "learning_rate": 7.683791340651337e-09,
+      "loss": 0.6898,
+      "step": 9712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5520360612107897,
+      "learning_rate": 7.555210788093847e-09,
+      "loss": 0.7203,
+      "step": 9713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.73846643821148,
+      "learning_rate": 7.427714759153937e-09,
+      "loss": 0.7294,
+      "step": 9714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7606361370390762,
+      "learning_rate": 7.301303267669424e-09,
+      "loss": 0.6608,
+      "step": 9715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8115631849090241,
+      "learning_rate": 7.1759763273604454e-09,
+      "loss": 0.814,
+      "step": 9716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6783185631212085,
+      "learning_rate": 7.051733951828343e-09,
+      "loss": 0.6538,
+      "step": 9717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.618824346005183,
+      "learning_rate": 6.928576154558997e-09,
+      "loss": 0.5846,
+      "step": 9718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8248486875496646,
+      "learning_rate": 6.806502948918381e-09,
+      "loss": 0.7861,
+      "step": 9719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6413599961753038,
+      "learning_rate": 6.685514348154787e-09,
+      "loss": 0.6407,
+      "step": 9720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8528942467367489,
+      "learning_rate": 6.565610365402153e-09,
+      "loss": 0.738,
+      "step": 9721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.818498691949131,
+      "learning_rate": 6.446791013671183e-09,
+      "loss": 0.7306,
+      "step": 9722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7186762257651678,
+      "learning_rate": 6.329056305860448e-09,
+      "loss": 0.6696,
+      "step": 9723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.757030635077287,
+      "learning_rate": 6.212406254746395e-09,
+      "loss": 0.7701,
+      "step": 9724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5623043797000618,
+      "learning_rate": 6.096840872991117e-09,
+      "loss": 0.662,
+      "step": 9725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5841618328268379,
+      "learning_rate": 5.9823601731356925e-09,
+      "loss": 0.748,
+      "step": 9726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6043392083126362,
+      "learning_rate": 5.86896416760685e-09,
+      "loss": 0.579,
+      "step": 9727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9265682270468332,
+      "learning_rate": 5.7566528687114095e-09,
+      "loss": 0.7247,
+      "step": 9728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.691440943069805,
+      "learning_rate": 5.645426288638511e-09,
+      "loss": 0.7415,
+      "step": 9729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6450176668324605,
+      "learning_rate": 5.53528443945961e-09,
+      "loss": 0.6279,
+      "step": 9730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7378168961948732,
+      "learning_rate": 5.426227333130696e-09,
+      "loss": 0.697,
+      "step": 9731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7252926169952139,
+      "learning_rate": 5.318254981486748e-09,
+      "loss": 0.6545,
+      "step": 9732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7953360616753136,
+      "learning_rate": 5.21136739624617e-09,
+      "loss": 0.7161,
+      "step": 9733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6194759118926911,
+      "learning_rate": 5.105564589011902e-09,
+      "loss": 0.677,
+      "step": 9734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7126316233812235,
+      "learning_rate": 5.000846571264762e-09,
+      "loss": 0.6515,
+      "step": 9735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9346810637478329,
+      "learning_rate": 4.897213354372321e-09,
+      "loss": 0.6876,
+      "step": 9736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.209066745028407,
+      "learning_rate": 4.7946649495811405e-09,
+      "loss": 0.7986,
+      "step": 9737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8773927683013005,
+      "learning_rate": 4.693201368021205e-09,
+      "loss": 0.7163,
+      "step": 9738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8109093998592674,
+      "learning_rate": 4.592822620705928e-09,
+      "loss": 0.7481,
+      "step": 9739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5717984737809685,
+      "learning_rate": 4.493528718528817e-09,
+      "loss": 0.6516,
+      "step": 9740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6172161663400688,
+      "learning_rate": 4.395319672266807e-09,
+      "loss": 0.6364,
+      "step": 9741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6682760972502344,
+      "learning_rate": 4.2981954925780385e-09,
+      "loss": 0.7554,
+      "step": 9742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.868953172176466,
+      "learning_rate": 4.202156190006301e-09,
+      "loss": 0.7247,
+      "step": 9743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7716773681131424,
+      "learning_rate": 4.1072017749732574e-09,
+      "loss": 0.6867,
+      "step": 9744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5398947349645191,
+      "learning_rate": 4.013332257785107e-09,
+      "loss": 0.5854,
+      "step": 9745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.52256409465704,
+      "learning_rate": 3.920547648630368e-09,
+      "loss": 0.6266,
+      "step": 9746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7581089180332012,
+      "learning_rate": 3.828847957577653e-09,
+      "loss": 0.6372,
+      "step": 9747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.596383318497569,
+      "learning_rate": 3.73823319458233e-09,
+      "loss": 0.6177,
+      "step": 9748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6520114837204025,
+      "learning_rate": 3.6487033694776466e-09,
+      "loss": 0.6509,
+      "step": 9749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7583325520760777,
+      "learning_rate": 3.560258491980273e-09,
+      "loss": 0.7194,
+      "step": 9750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.596475826131414,
+      "learning_rate": 3.4728985716903083e-09,
+      "loss": 0.7752,
+      "step": 9751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.436844800858631,
+      "learning_rate": 3.3866236180879476e-09,
+      "loss": 0.6588,
+      "step": 9752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8173776459488546,
+      "learning_rate": 3.3014336405390313e-09,
+      "loss": 0.7334,
+      "step": 9753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7236920470298163,
+      "learning_rate": 3.2173286482883868e-09,
+      "loss": 0.7585,
+      "step": 9754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6707557679087048,
+      "learning_rate": 3.1343086504653785e-09,
+      "loss": 0.6727,
+      "step": 9755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.559160997871232,
+      "learning_rate": 3.0523736560783558e-09,
+      "loss": 0.6343,
+      "step": 9756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6592056636731305,
+      "learning_rate": 2.971523674022425e-09,
+      "loss": 0.6621,
+      "step": 9757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7959185329193494,
+      "learning_rate": 2.8917587130705695e-09,
+      "loss": 0.7023,
+      "step": 9758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5316612752649197,
+      "learning_rate": 2.8130787818814177e-09,
+      "loss": 0.7049,
+      "step": 9759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5013465330785167,
+      "learning_rate": 2.7354838889948055e-09,
+      "loss": 0.676,
+      "step": 9760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6680553354247996,
+      "learning_rate": 2.6589740428306644e-09,
+      "loss": 0.7261,
+      "step": 9761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5739749500286022,
+      "learning_rate": 2.5835492516945725e-09,
+      "loss": 0.5215,
+      "step": 9762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.80329783598832,
+      "learning_rate": 2.5092095237722048e-09,
+      "loss": 0.7339,
+      "step": 9763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7697361499388937,
+      "learning_rate": 2.4359548671315515e-09,
+      "loss": 0.783,
+      "step": 9764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8060799010027864,
+      "learning_rate": 2.36378528972292e-09,
+      "loss": 0.6659,
+      "step": 9765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5320767192304863,
+      "learning_rate": 2.2927007993811535e-09,
+      "loss": 0.6163,
+      "step": 9766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8341164287247476,
+      "learning_rate": 2.222701403818972e-09,
+      "loss": 0.7484,
+      "step": 9767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6898445181700266,
+      "learning_rate": 2.153787110634742e-09,
+      "loss": 0.7118,
+      "step": 9768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.73467212577717,
+      "learning_rate": 2.0859579273091457e-09,
+      "loss": 0.6583,
+      "step": 9769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8342421843135626,
+      "learning_rate": 2.019213861201852e-09,
+      "loss": 0.8027,
+      "step": 9770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7248601263057854,
+      "learning_rate": 1.953554919559286e-09,
+      "loss": 0.7182,
+      "step": 9771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7852646128943321,
+      "learning_rate": 1.8889811095046396e-09,
+      "loss": 0.6439,
+      "step": 9772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.797497617117918,
+      "learning_rate": 1.8254924380489702e-09,
+      "loss": 0.7096,
+      "step": 9773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6025539122176873,
+      "learning_rate": 1.7630889120823224e-09,
+      "loss": 0.6443,
+      "step": 9774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5688322495059948,
+      "learning_rate": 1.7017705383781668e-09,
+      "loss": 0.7314,
+      "step": 9775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7617953713834467,
+      "learning_rate": 1.64153732359007e-09,
+      "loss": 0.6941,
+      "step": 9776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8437779372800862,
+      "learning_rate": 1.5823892742561352e-09,
+      "loss": 0.6198,
+      "step": 9777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5327724689221702,
+      "learning_rate": 1.5243263967956722e-09,
+      "loss": 0.7511,
+      "step": 9778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.632644536650647,
+      "learning_rate": 1.467348697511417e-09,
+      "loss": 0.7554,
+      "step": 9779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8949552888377248,
+      "learning_rate": 1.411456182587312e-09,
+      "loss": 0.7169,
+      "step": 9780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6767940261577814,
+      "learning_rate": 1.356648858088505e-09,
+      "loss": 0.6894,
+      "step": 9781
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7349273531608402,
+      "learning_rate": 1.302926729964682e-09,
+      "loss": 0.6867,
+      "step": 9782
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6491775044986772,
+      "learning_rate": 1.2502898040456235e-09,
+      "loss": 0.6245,
+      "step": 9783
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5181692314067778,
+      "learning_rate": 1.1987380860456477e-09,
+      "loss": 0.6693,
+      "step": 9784
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5327161046836033,
+      "learning_rate": 1.148271581558058e-09,
+      "loss": 0.6834,
+      "step": 9785
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6414243607298673,
+      "learning_rate": 1.098890296060695e-09,
+      "loss": 0.6154,
+      "step": 9786
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.844885336324718,
+      "learning_rate": 1.0505942349137155e-09,
+      "loss": 0.8566,
+      "step": 9787
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5735819481588773,
+      "learning_rate": 1.003383403358482e-09,
+      "loss": 0.7284,
+      "step": 9788
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7929415551134282,
+      "learning_rate": 9.572578065197846e-10,
+      "loss": 0.6794,
+      "step": 9789
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8147303747186208,
+      "learning_rate": 9.122174494025082e-10,
+      "loss": 0.6762,
+      "step": 9790
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8998667504295066,
+      "learning_rate": 8.682623368971854e-10,
+      "loss": 0.7023,
+      "step": 9791
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5999912018017683,
+      "learning_rate": 8.253924737711139e-10,
+      "loss": 0.6024,
+      "step": 9792
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6802611122367843,
+      "learning_rate": 7.83607864680569e-10,
+      "loss": 0.677,
+      "step": 9793
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9299673966168387,
+      "learning_rate": 7.429085141585912e-10,
+      "loss": 0.6575,
+      "step": 9794
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8079801054976152,
+      "learning_rate": 7.032944266227582e-10,
+      "loss": 0.7263,
+      "step": 9795
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7053584530993402,
+      "learning_rate": 6.647656063729635e-10,
+      "loss": 0.7107,
+      "step": 9796
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.752559768160014,
+      "learning_rate": 6.273220575914174e-10,
+      "loss": 0.6197,
+      "step": 9797
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6022948579476681,
+      "learning_rate": 5.909637843404259e-10,
+      "loss": 0.7039,
+      "step": 9798
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6826875055361312,
+      "learning_rate": 5.556907905679421e-10,
+      "loss": 0.6946,
+      "step": 9799
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6338947651630946,
+      "learning_rate": 5.215030801009046e-10,
+      "loss": 0.7023,
+      "step": 9800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.669612829015242,
+      "learning_rate": 4.884006566496791e-10,
+      "loss": 0.7131,
+      "step": 9801
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.595339314335866,
+      "learning_rate": 4.5638352380805716e-10,
+      "loss": 0.5837,
+      "step": 9802
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7544681428451936,
+      "learning_rate": 4.2545168505103707e-10,
+      "loss": 0.6499,
+      "step": 9803
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6948058791459002,
+      "learning_rate": 3.9560514373593317e-10,
+      "loss": 0.6961,
+      "step": 9804
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7957334177664486,
+      "learning_rate": 3.6684390310015584e-10,
+      "loss": 0.7444,
+      "step": 9805
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5597201209989984,
+      "learning_rate": 3.391679662678726e-10,
+      "loss": 0.6784,
+      "step": 9806
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8096052129609148,
+      "learning_rate": 3.1257733624112665e-10,
+      "loss": 0.6936,
+      "step": 9807
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6849798366893498,
+      "learning_rate": 2.8707201590649767e-10,
+      "loss": 0.6742,
+      "step": 9808
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.838312020306633,
+      "learning_rate": 2.6265200803177184e-10,
+      "loss": 0.7065,
+      "step": 9809
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.767900214133964,
+      "learning_rate": 2.3931731526927183e-10,
+      "loss": 0.8026,
+      "step": 9810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5904778245566455,
+      "learning_rate": 2.1706794014919596e-10,
+      "loss": 0.7111,
+      "step": 9811
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5468126215436857,
+      "learning_rate": 1.9590388508627934e-10,
+      "loss": 0.7569,
+      "step": 9812
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5375661558070106,
+      "learning_rate": 1.758251523797938e-10,
+      "loss": 0.611,
+      "step": 9813
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8357286117999407,
+      "learning_rate": 1.5683174420799696e-10,
+      "loss": 0.7772,
+      "step": 9814
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1584454747868387,
+      "learning_rate": 1.3892366263146273e-10,
+      "loss": 0.77,
+      "step": 9815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.674939245192305,
+      "learning_rate": 1.2210090959419162e-10,
+      "loss": 0.609,
+      "step": 9816
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5174189719877236,
+      "learning_rate": 1.0636348692250054e-10,
+      "loss": 0.585,
+      "step": 9817
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6701519955880866,
+      "learning_rate": 9.171139632391245e-11,
+      "loss": 0.748,
+      "step": 9818
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5132318407758787,
+      "learning_rate": 7.8144639389377e-11,
+      "loss": 0.5916,
+      "step": 9819
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6711317998681496,
+      "learning_rate": 6.566321759104987e-11,
+      "loss": 0.7035,
+      "step": 9820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7089305325553747,
+      "learning_rate": 5.426713228340319e-11,
+      "loss": 0.6716,
+      "step": 9821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7533657894024617,
+      "learning_rate": 4.3956384703225384e-11,
+      "loss": 0.7033,
+      "step": 9822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7866771307941718,
+      "learning_rate": 3.4730975969621275e-11,
+      "loss": 0.7179,
+      "step": 9823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7858168287448422,
+      "learning_rate": 2.659090708401202e-11,
+      "loss": 0.6644,
+      "step": 9824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.849749275071706,
+      "learning_rate": 1.9536178930135152e-11,
+      "loss": 0.586,
+      "step": 9825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.779817554106832,
+      "learning_rate": 1.356679227404456e-11,
+      "loss": 0.6623,
+      "step": 9826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7365065100153105,
+      "learning_rate": 8.682747761890042e-12,
+      "loss": 0.6541,
+      "step": 9827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7159852243362201,
+      "learning_rate": 4.884045925468428e-12,
+      "loss": 0.6903,
+      "step": 9828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9049278490404138,
+      "learning_rate": 2.170687175562236e-12,
+      "loss": 0.5288,
+      "step": 9829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6298596172428135,
+      "learning_rate": 5.426718086010141e-13,
+      "loss": 0.7138,
+      "step": 9830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6170033821053835,
+      "learning_rate": 0.0,
+      "loss": 0.7636,
+      "step": 9831
+    },
+    {
+      "epoch": 1.0,
+      "step": 9831,
+      "total_flos": 1617537550008320.0,
+      "train_loss": 0.7271600768035261,
+      "train_runtime": 125510.4499,
+      "train_samples_per_second": 10.025,
+      "train_steps_per_second": 0.078
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9831,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1617537550008320.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}