diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6678 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.993684210526316,
+  "eval_steps": 500,
+  "global_step": 948,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003157894736842105,
+      "grad_norm": 6.864107608795166,
+      "learning_rate": 1.0526315789473685e-07,
+      "loss": 1.2166,
+      "step": 1
+    },
+    {
+      "epoch": 0.00631578947368421,
+      "grad_norm": 7.149033546447754,
+      "learning_rate": 2.105263157894737e-07,
+      "loss": 1.2292,
+      "step": 2
+    },
+    {
+      "epoch": 0.009473684210526316,
+      "grad_norm": 6.731090545654297,
+      "learning_rate": 3.1578947368421055e-07,
+      "loss": 1.1835,
+      "step": 3
+    },
+    {
+      "epoch": 0.01263157894736842,
+      "grad_norm": 7.016407012939453,
+      "learning_rate": 4.210526315789474e-07,
+      "loss": 1.2222,
+      "step": 4
+    },
+    {
+      "epoch": 0.015789473684210527,
+      "grad_norm": 6.585211753845215,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 1.2031,
+      "step": 5
+    },
+    {
+      "epoch": 0.018947368421052633,
+      "grad_norm": 6.720523834228516,
+      "learning_rate": 6.315789473684211e-07,
+      "loss": 1.1972,
+      "step": 6
+    },
+    {
+      "epoch": 0.022105263157894735,
+      "grad_norm": 6.992764472961426,
+      "learning_rate": 7.368421052631579e-07,
+      "loss": 1.1867,
+      "step": 7
+    },
+    {
+      "epoch": 0.02526315789473684,
+      "grad_norm": 6.5448174476623535,
+      "learning_rate": 8.421052631578948e-07,
+      "loss": 1.2113,
+      "step": 8
+    },
+    {
+      "epoch": 0.028421052631578948,
+      "grad_norm": 6.110010147094727,
+      "learning_rate": 9.473684210526317e-07,
+      "loss": 1.1666,
+      "step": 9
+    },
+    {
+      "epoch": 0.031578947368421054,
+      "grad_norm": 6.425546169281006,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 1.1722,
+      "step": 10
+    },
+    {
+      "epoch": 0.034736842105263156,
+      "grad_norm": 4.949359893798828,
+      "learning_rate": 1.1578947368421053e-06,
+      "loss": 1.1464,
+      "step": 11
+    },
+    {
+      "epoch": 0.037894736842105266,
+      "grad_norm": 4.857133865356445,
+      "learning_rate": 1.2631578947368422e-06,
+      "loss": 1.1495,
+      "step": 12
+    },
+    {
+      "epoch": 0.04105263157894737,
+      "grad_norm": 4.663426876068115,
+      "learning_rate": 1.3684210526315791e-06,
+      "loss": 1.1108,
+      "step": 13
+    },
+    {
+      "epoch": 0.04421052631578947,
+      "grad_norm": 2.899270534515381,
+      "learning_rate": 1.4736842105263159e-06,
+      "loss": 1.0779,
+      "step": 14
+    },
+    {
+      "epoch": 0.04736842105263158,
+      "grad_norm": 2.799199104309082,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 1.0596,
+      "step": 15
+    },
+    {
+      "epoch": 0.05052631578947368,
+      "grad_norm": 2.7462809085845947,
+      "learning_rate": 1.6842105263157895e-06,
+      "loss": 1.0846,
+      "step": 16
+    },
+    {
+      "epoch": 0.05368421052631579,
+      "grad_norm": 2.6440353393554688,
+      "learning_rate": 1.7894736842105265e-06,
+      "loss": 1.0408,
+      "step": 17
+    },
+    {
+      "epoch": 0.056842105263157895,
+      "grad_norm": 2.496100664138794,
+      "learning_rate": 1.8947368421052634e-06,
+      "loss": 1.0447,
+      "step": 18
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5369954109191895,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.0338,
+      "step": 19
+    },
+    {
+      "epoch": 0.06315789473684211,
+      "grad_norm": 3.36142635345459,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 1.0461,
+      "step": 20
+    },
+    {
+      "epoch": 0.06631578947368422,
+      "grad_norm": 3.2484495639801025,
+      "learning_rate": 2.2105263157894738e-06,
+      "loss": 1.0202,
+      "step": 21
+    },
+    {
+      "epoch": 0.06947368421052631,
+      "grad_norm": 3.0597786903381348,
+      "learning_rate": 2.3157894736842105e-06,
+      "loss": 1.0088,
+      "step": 22
+    },
+    {
+      "epoch": 0.07263157894736842,
+      "grad_norm": 2.594919204711914,
+      "learning_rate": 2.4210526315789477e-06,
+      "loss": 0.9878,
+      "step": 23
+    },
+    {
+      "epoch": 0.07578947368421053,
+      "grad_norm": 2.3219525814056396,
+      "learning_rate": 2.5263157894736844e-06,
+      "loss": 1.0149,
+      "step": 24
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 1.812867283821106,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.9818,
+      "step": 25
+    },
+    {
+      "epoch": 0.08210526315789474,
+      "grad_norm": 1.581230640411377,
+      "learning_rate": 2.7368421052631583e-06,
+      "loss": 0.9757,
+      "step": 26
+    },
+    {
+      "epoch": 0.08526315789473685,
+      "grad_norm": 1.8027225732803345,
+      "learning_rate": 2.842105263157895e-06,
+      "loss": 0.9676,
+      "step": 27
+    },
+    {
+      "epoch": 0.08842105263157894,
+      "grad_norm": 1.7598600387573242,
+      "learning_rate": 2.9473684210526317e-06,
+      "loss": 0.9181,
+      "step": 28
+    },
+    {
+      "epoch": 0.09157894736842105,
+      "grad_norm": 1.7967835664749146,
+      "learning_rate": 3.052631578947369e-06,
+      "loss": 0.9336,
+      "step": 29
+    },
+    {
+      "epoch": 0.09473684210526316,
+      "grad_norm": 1.3898029327392578,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.9156,
+      "step": 30
+    },
+    {
+      "epoch": 0.09789473684210526,
+      "grad_norm": 1.2304682731628418,
+      "learning_rate": 3.2631578947368423e-06,
+      "loss": 0.936,
+      "step": 31
+    },
+    {
+      "epoch": 0.10105263157894737,
+      "grad_norm": 1.1240628957748413,
+      "learning_rate": 3.368421052631579e-06,
+      "loss": 0.8984,
+      "step": 32
+    },
+    {
+      "epoch": 0.10421052631578948,
+      "grad_norm": 1.1114623546600342,
+      "learning_rate": 3.473684210526316e-06,
+      "loss": 0.9032,
+      "step": 33
+    },
+    {
+      "epoch": 0.10736842105263159,
+      "grad_norm": 1.2093194723129272,
+      "learning_rate": 3.578947368421053e-06,
+      "loss": 0.9466,
+      "step": 34
+    },
+    {
+      "epoch": 0.11052631578947368,
+      "grad_norm": 1.1614065170288086,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 0.8888,
+      "step": 35
+    },
+    {
+      "epoch": 0.11368421052631579,
+      "grad_norm": 0.9620938897132874,
+      "learning_rate": 3.789473684210527e-06,
+      "loss": 0.8985,
+      "step": 36
+    },
+    {
+      "epoch": 0.1168421052631579,
+      "grad_norm": 0.8165697455406189,
+      "learning_rate": 3.894736842105263e-06,
+      "loss": 0.8922,
+      "step": 37
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9624667763710022,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.9059,
+      "step": 38
+    },
+    {
+      "epoch": 0.1231578947368421,
+      "grad_norm": 0.9526475667953491,
+      "learning_rate": 4.105263157894737e-06,
+      "loss": 0.8697,
+      "step": 39
+    },
+    {
+      "epoch": 0.12631578947368421,
+      "grad_norm": 0.9052897691726685,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.8752,
+      "step": 40
+    },
+    {
+      "epoch": 0.12947368421052632,
+      "grad_norm": 0.8724302053451538,
+      "learning_rate": 4.315789473684211e-06,
+      "loss": 0.8745,
+      "step": 41
+    },
+    {
+      "epoch": 0.13263157894736843,
+      "grad_norm": 0.8744531273841858,
+      "learning_rate": 4.4210526315789476e-06,
+      "loss": 0.833,
+      "step": 42
+    },
+    {
+      "epoch": 0.13578947368421052,
+      "grad_norm": 0.8191508650779724,
+      "learning_rate": 4.526315789473685e-06,
+      "loss": 0.8681,
+      "step": 43
+    },
+    {
+      "epoch": 0.13894736842105262,
+      "grad_norm": 0.8925843238830566,
+      "learning_rate": 4.631578947368421e-06,
+      "loss": 0.8522,
+      "step": 44
+    },
+    {
+      "epoch": 0.14210526315789473,
+      "grad_norm": 0.89957195520401,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.8732,
+      "step": 45
+    },
+    {
+      "epoch": 0.14526315789473684,
+      "grad_norm": 0.8851773738861084,
+      "learning_rate": 4.842105263157895e-06,
+      "loss": 0.8991,
+      "step": 46
+    },
+    {
+      "epoch": 0.14842105263157895,
+      "grad_norm": 0.7817597985267639,
+      "learning_rate": 4.947368421052632e-06,
+      "loss": 0.8711,
+      "step": 47
+    },
+    {
+      "epoch": 0.15157894736842106,
+      "grad_norm": 0.7551989555358887,
+      "learning_rate": 5.052631578947369e-06,
+      "loss": 0.8306,
+      "step": 48
+    },
+    {
+      "epoch": 0.15473684210526314,
+      "grad_norm": 0.8163503408432007,
+      "learning_rate": 5.157894736842106e-06,
+      "loss": 0.8403,
+      "step": 49
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.7936963438987732,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.837,
+      "step": 50
+    },
+    {
+      "epoch": 0.16105263157894736,
+      "grad_norm": 0.7440241575241089,
+      "learning_rate": 5.36842105263158e-06,
+      "loss": 0.8538,
+      "step": 51
+    },
+    {
+      "epoch": 0.16421052631578947,
+      "grad_norm": 0.7333534359931946,
+      "learning_rate": 5.4736842105263165e-06,
+      "loss": 0.819,
+      "step": 52
+    },
+    {
+      "epoch": 0.16736842105263158,
+      "grad_norm": 0.6583252549171448,
+      "learning_rate": 5.578947368421052e-06,
+      "loss": 0.7879,
+      "step": 53
+    },
+    {
+      "epoch": 0.1705263157894737,
+      "grad_norm": 0.684806764125824,
+      "learning_rate": 5.68421052631579e-06,
+      "loss": 0.8325,
+      "step": 54
+    },
+    {
+      "epoch": 0.1736842105263158,
+      "grad_norm": 0.7598180174827576,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 0.8341,
+      "step": 55
+    },
+    {
+      "epoch": 0.17684210526315788,
+      "grad_norm": 0.8507996201515198,
+      "learning_rate": 5.8947368421052634e-06,
+      "loss": 0.8428,
+      "step": 56
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7600204348564148,
+      "learning_rate": 6e-06,
+      "loss": 0.818,
+      "step": 57
+    },
+    {
+      "epoch": 0.1831578947368421,
+      "grad_norm": 0.6492787599563599,
+      "learning_rate": 6.105263157894738e-06,
+      "loss": 0.8346,
+      "step": 58
+    },
+    {
+      "epoch": 0.1863157894736842,
+      "grad_norm": 0.6897916793823242,
+      "learning_rate": 6.2105263157894745e-06,
+      "loss": 0.8409,
+      "step": 59
+    },
+    {
+      "epoch": 0.18947368421052632,
+      "grad_norm": 0.7259018421173096,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 0.8285,
+      "step": 60
+    },
+    {
+      "epoch": 0.19263157894736843,
+      "grad_norm": 0.8188095688819885,
+      "learning_rate": 6.421052631578948e-06,
+      "loss": 0.8816,
+      "step": 61
+    },
+    {
+      "epoch": 0.1957894736842105,
+      "grad_norm": 0.6784875988960266,
+      "learning_rate": 6.526315789473685e-06,
+      "loss": 0.8153,
+      "step": 62
+    },
+    {
+      "epoch": 0.19894736842105262,
+      "grad_norm": 0.740426242351532,
+      "learning_rate": 6.631578947368421e-06,
+      "loss": 0.8243,
+      "step": 63
+    },
+    {
+      "epoch": 0.20210526315789473,
+      "grad_norm": 0.7158690094947815,
+      "learning_rate": 6.736842105263158e-06,
+      "loss": 0.828,
+      "step": 64
+    },
+    {
+      "epoch": 0.20526315789473684,
+      "grad_norm": 0.7803067564964294,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.7897,
+      "step": 65
+    },
+    {
+      "epoch": 0.20842105263157895,
+      "grad_norm": 0.6472960710525513,
+      "learning_rate": 6.947368421052632e-06,
+      "loss": 0.7908,
+      "step": 66
+    },
+    {
+      "epoch": 0.21157894736842106,
+      "grad_norm": 0.7025761008262634,
+      "learning_rate": 7.052631578947369e-06,
+      "loss": 0.8408,
+      "step": 67
+    },
+    {
+      "epoch": 0.21473684210526317,
+      "grad_norm": 0.7511818408966064,
+      "learning_rate": 7.157894736842106e-06,
+      "loss": 0.7913,
+      "step": 68
+    },
+    {
+      "epoch": 0.21789473684210525,
+      "grad_norm": 0.6920287609100342,
+      "learning_rate": 7.263157894736843e-06,
+      "loss": 0.8173,
+      "step": 69
+    },
+    {
+      "epoch": 0.22105263157894736,
+      "grad_norm": 0.6897710561752319,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.8213,
+      "step": 70
+    },
+    {
+      "epoch": 0.22421052631578947,
+      "grad_norm": 0.7388243675231934,
+      "learning_rate": 7.473684210526316e-06,
+      "loss": 0.8039,
+      "step": 71
+    },
+    {
+      "epoch": 0.22736842105263158,
+      "grad_norm": 0.8480104804039001,
+      "learning_rate": 7.578947368421054e-06,
+      "loss": 0.8235,
+      "step": 72
+    },
+    {
+      "epoch": 0.2305263157894737,
+      "grad_norm": 0.7924239635467529,
+      "learning_rate": 7.68421052631579e-06,
+      "loss": 0.8256,
+      "step": 73
+    },
+    {
+      "epoch": 0.2336842105263158,
+      "grad_norm": 0.7806857228279114,
+      "learning_rate": 7.789473684210526e-06,
+      "loss": 0.8284,
+      "step": 74
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 0.7032557129859924,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.7895,
+      "step": 75
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8388243913650513,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.8032,
+      "step": 76
+    },
+    {
+      "epoch": 0.2431578947368421,
+      "grad_norm": 0.8523675203323364,
+      "learning_rate": 8.105263157894736e-06,
+      "loss": 0.8118,
+      "step": 77
+    },
+    {
+      "epoch": 0.2463157894736842,
+      "grad_norm": 0.7454504370689392,
+      "learning_rate": 8.210526315789475e-06,
+      "loss": 0.7955,
+      "step": 78
+    },
+    {
+      "epoch": 0.24947368421052632,
+      "grad_norm": 0.737902820110321,
+      "learning_rate": 8.315789473684212e-06,
+      "loss": 0.8337,
+      "step": 79
+    },
+    {
+      "epoch": 0.25263157894736843,
+      "grad_norm": 0.7588112950325012,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.7851,
+      "step": 80
+    },
+    {
+      "epoch": 0.2557894736842105,
+      "grad_norm": 0.812930166721344,
+      "learning_rate": 8.526315789473685e-06,
+      "loss": 0.8024,
+      "step": 81
+    },
+    {
+      "epoch": 0.25894736842105265,
+      "grad_norm": 0.9560610055923462,
+      "learning_rate": 8.631578947368422e-06,
+      "loss": 0.8093,
+      "step": 82
+    },
+    {
+      "epoch": 0.26210526315789473,
+      "grad_norm": 0.9282992482185364,
+      "learning_rate": 8.736842105263158e-06,
+      "loss": 0.8169,
+      "step": 83
+    },
+    {
+      "epoch": 0.26526315789473687,
+      "grad_norm": 0.7830009460449219,
+      "learning_rate": 8.842105263157895e-06,
+      "loss": 0.8062,
+      "step": 84
+    },
+    {
+      "epoch": 0.26842105263157895,
+      "grad_norm": 0.8576416969299316,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.8023,
+      "step": 85
+    },
+    {
+      "epoch": 0.27157894736842103,
+      "grad_norm": 0.8227952122688293,
+      "learning_rate": 9.05263157894737e-06,
+      "loss": 0.805,
+      "step": 86
+    },
+    {
+      "epoch": 0.27473684210526317,
+      "grad_norm": 0.8914945721626282,
+      "learning_rate": 9.157894736842105e-06,
+      "loss": 0.8269,
+      "step": 87
+    },
+    {
+      "epoch": 0.27789473684210525,
+      "grad_norm": 0.81199711561203,
+      "learning_rate": 9.263157894736842e-06,
+      "loss": 0.8056,
+      "step": 88
+    },
+    {
+      "epoch": 0.2810526315789474,
+      "grad_norm": 1.0146549940109253,
+      "learning_rate": 9.36842105263158e-06,
+      "loss": 0.7905,
+      "step": 89
+    },
+    {
+      "epoch": 0.28421052631578947,
+      "grad_norm": 0.8971161246299744,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.8079,
+      "step": 90
+    },
+    {
+      "epoch": 0.2873684210526316,
+      "grad_norm": 0.9439178705215454,
+      "learning_rate": 9.578947368421054e-06,
+      "loss": 0.8126,
+      "step": 91
+    },
+    {
+      "epoch": 0.2905263157894737,
+      "grad_norm": 0.9185954332351685,
+      "learning_rate": 9.68421052631579e-06,
+      "loss": 0.7945,
+      "step": 92
+    },
+    {
+      "epoch": 0.29368421052631577,
+      "grad_norm": 0.7818227410316467,
+      "learning_rate": 9.789473684210527e-06,
+      "loss": 0.7825,
+      "step": 93
+    },
+    {
+      "epoch": 0.2968421052631579,
+      "grad_norm": 0.7514786124229431,
+      "learning_rate": 9.894736842105264e-06,
+      "loss": 0.8049,
+      "step": 94
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0437406301498413,
+      "learning_rate": 1e-05,
+      "loss": 0.8204,
+      "step": 95
+    },
+    {
+      "epoch": 0.3031578947368421,
+      "grad_norm": 0.921838104724884,
+      "learning_rate": 9.999966088952842e-06,
+      "loss": 0.8107,
+      "step": 96
+    },
+    {
+      "epoch": 0.3063157894736842,
+      "grad_norm": 0.7265084981918335,
+      "learning_rate": 9.99986435627135e-06,
+      "loss": 0.8067,
+      "step": 97
+    },
+    {
+      "epoch": 0.3094736842105263,
+      "grad_norm": 1.0714375972747803,
+      "learning_rate": 9.999694803335468e-06,
+      "loss": 0.8052,
+      "step": 98
+    },
+    {
+      "epoch": 0.3126315789473684,
+      "grad_norm": 0.9621460437774658,
+      "learning_rate": 9.999457432445087e-06,
+      "loss": 0.8122,
+      "step": 99
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.7225667834281921,
+      "learning_rate": 9.999152246820001e-06,
+      "loss": 0.8104,
+      "step": 100
+    },
+    {
+      "epoch": 0.31894736842105265,
+      "grad_norm": 0.8660763502120972,
+      "learning_rate": 9.998779250599877e-06,
+      "loss": 0.8005,
+      "step": 101
+    },
+    {
+      "epoch": 0.32210526315789473,
+      "grad_norm": 0.951799213886261,
+      "learning_rate": 9.998338448844193e-06,
+      "loss": 0.8022,
+      "step": 102
+    },
+    {
+      "epoch": 0.32526315789473687,
+      "grad_norm": 0.8598304986953735,
+      "learning_rate": 9.997829847532165e-06,
+      "loss": 0.7862,
+      "step": 103
+    },
+    {
+      "epoch": 0.32842105263157895,
+      "grad_norm": 0.7481747269630432,
+      "learning_rate": 9.99725345356268e-06,
+      "loss": 0.7933,
+      "step": 104
+    },
+    {
+      "epoch": 0.33157894736842103,
+      "grad_norm": 0.9335681200027466,
+      "learning_rate": 9.996609274754183e-06,
+      "loss": 0.7822,
+      "step": 105
+    },
+    {
+      "epoch": 0.33473684210526317,
+      "grad_norm": 0.8390291929244995,
+      "learning_rate": 9.995897319844588e-06,
+      "loss": 0.7838,
+      "step": 106
+    },
+    {
+      "epoch": 0.33789473684210525,
+      "grad_norm": 0.7386859655380249,
+      "learning_rate": 9.995117598491146e-06,
+      "loss": 0.7872,
+      "step": 107
+    },
+    {
+      "epoch": 0.3410526315789474,
+      "grad_norm": 0.8380564451217651,
+      "learning_rate": 9.994270121270327e-06,
+      "loss": 0.7568,
+      "step": 108
+    },
+    {
+      "epoch": 0.34421052631578947,
+      "grad_norm": 0.7734573483467102,
+      "learning_rate": 9.993354899677665e-06,
+      "loss": 0.8004,
+      "step": 109
+    },
+    {
+      "epoch": 0.3473684210526316,
+      "grad_norm": 0.8333052396774292,
+      "learning_rate": 9.99237194612761e-06,
+      "loss": 0.7686,
+      "step": 110
+    },
+    {
+      "epoch": 0.3505263157894737,
+      "grad_norm": 0.9416193962097168,
+      "learning_rate": 9.991321273953357e-06,
+      "loss": 0.7973,
+      "step": 111
+    },
+    {
+      "epoch": 0.35368421052631577,
+      "grad_norm": 0.7098072171211243,
+      "learning_rate": 9.99020289740666e-06,
+      "loss": 0.7547,
+      "step": 112
+    },
+    {
+      "epoch": 0.3568421052631579,
+      "grad_norm": 0.9404763579368591,
+      "learning_rate": 9.989016831657652e-06,
+      "loss": 0.8093,
+      "step": 113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0210902690887451,
+      "learning_rate": 9.987763092794621e-06,
+      "loss": 0.7939,
+      "step": 114
+    },
+    {
+      "epoch": 0.3631578947368421,
+      "grad_norm": 0.7902900576591492,
+      "learning_rate": 9.986441697823808e-06,
+      "loss": 0.7889,
+      "step": 115
+    },
+    {
+      "epoch": 0.3663157894736842,
+      "grad_norm": 0.9824138283729553,
+      "learning_rate": 9.985052664669168e-06,
+      "loss": 0.7888,
+      "step": 116
+    },
+    {
+      "epoch": 0.36947368421052634,
+      "grad_norm": 0.9166287183761597,
+      "learning_rate": 9.983596012172127e-06,
+      "loss": 0.7845,
+      "step": 117
+    },
+    {
+      "epoch": 0.3726315789473684,
+      "grad_norm": 0.7522174715995789,
+      "learning_rate": 9.982071760091334e-06,
+      "loss": 0.7619,
+      "step": 118
+    },
+    {
+      "epoch": 0.3757894736842105,
+      "grad_norm": 0.9811515212059021,
+      "learning_rate": 9.980479929102377e-06,
+      "loss": 0.7943,
+      "step": 119
+    },
+    {
+      "epoch": 0.37894736842105264,
+      "grad_norm": 0.7924212217330933,
+      "learning_rate": 9.978820540797521e-06,
+      "loss": 0.7928,
+      "step": 120
+    },
+    {
+      "epoch": 0.3821052631578947,
+      "grad_norm": 0.8440699577331543,
+      "learning_rate": 9.977093617685404e-06,
+      "loss": 0.8005,
+      "step": 121
+    },
+    {
+      "epoch": 0.38526315789473686,
+      "grad_norm": 0.7170959115028381,
+      "learning_rate": 9.975299183190734e-06,
+      "loss": 0.7844,
+      "step": 122
+    },
+    {
+      "epoch": 0.38842105263157894,
+      "grad_norm": 0.9932653307914734,
+      "learning_rate": 9.973437261653973e-06,
+      "loss": 0.7932,
+      "step": 123
+    },
+    {
+      "epoch": 0.391578947368421,
+      "grad_norm": 0.7866117358207703,
+      "learning_rate": 9.971507878331005e-06,
+      "loss": 0.7672,
+      "step": 124
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 0.791227400302887,
+      "learning_rate": 9.96951105939279e-06,
+      "loss": 0.7678,
+      "step": 125
+    },
+    {
+      "epoch": 0.39789473684210525,
+      "grad_norm": 1.0364530086517334,
+      "learning_rate": 9.96744683192502e-06,
+      "loss": 0.7814,
+      "step": 126
+    },
+    {
+      "epoch": 0.4010526315789474,
+      "grad_norm": 0.6801534295082092,
+      "learning_rate": 9.96531522392774e-06,
+      "loss": 0.751,
+      "step": 127
+    },
+    {
+      "epoch": 0.40421052631578946,
+      "grad_norm": 1.0359455347061157,
+      "learning_rate": 9.963116264314974e-06,
+      "loss": 0.7701,
+      "step": 128
+    },
+    {
+      "epoch": 0.4073684210526316,
+      "grad_norm": 0.8113856315612793,
+      "learning_rate": 9.960849982914332e-06,
+      "loss": 0.7924,
+      "step": 129
+    },
+    {
+      "epoch": 0.4105263157894737,
+      "grad_norm": 0.7722939848899841,
+      "learning_rate": 9.958516410466601e-06,
+      "loss": 0.7891,
+      "step": 130
+    },
+    {
+      "epoch": 0.41368421052631577,
+      "grad_norm": 0.8695185780525208,
+      "learning_rate": 9.956115578625339e-06,
+      "loss": 0.7795,
+      "step": 131
+    },
+    {
+      "epoch": 0.4168421052631579,
+      "grad_norm": 0.6542952060699463,
+      "learning_rate": 9.953647519956432e-06,
+      "loss": 0.7518,
+      "step": 132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7686191201210022,
+      "learning_rate": 9.951112267937663e-06,
+      "loss": 0.7851,
+      "step": 133
+    },
+    {
+      "epoch": 0.4231578947368421,
+      "grad_norm": 0.7790025472640991,
+      "learning_rate": 9.948509856958253e-06,
+      "loss": 0.7453,
+      "step": 134
+    },
+    {
+      "epoch": 0.4263157894736842,
+      "grad_norm": 0.7110816240310669,
+      "learning_rate": 9.945840322318391e-06,
+      "loss": 0.8121,
+      "step": 135
+    },
+    {
+      "epoch": 0.42947368421052634,
+      "grad_norm": 0.7633756995201111,
+      "learning_rate": 9.943103700228768e-06,
+      "loss": 0.7982,
+      "step": 136
+    },
+    {
+      "epoch": 0.4326315789473684,
+      "grad_norm": 0.7634323239326477,
+      "learning_rate": 9.940300027810067e-06,
+      "loss": 0.8097,
+      "step": 137
+    },
+    {
+      "epoch": 0.4357894736842105,
+      "grad_norm": 0.7589544057846069,
+      "learning_rate": 9.93742934309248e-06,
+      "loss": 0.7697,
+      "step": 138
+    },
+    {
+      "epoch": 0.43894736842105264,
+      "grad_norm": 0.820608377456665,
+      "learning_rate": 9.934491685015173e-06,
+      "loss": 0.7917,
+      "step": 139
+    },
+    {
+      "epoch": 0.4421052631578947,
+      "grad_norm": 0.801396906375885,
+      "learning_rate": 9.931487093425775e-06,
+      "loss": 0.7908,
+      "step": 140
+    },
+    {
+      "epoch": 0.44526315789473686,
+      "grad_norm": 0.9706858992576599,
+      "learning_rate": 9.928415609079821e-06,
+      "loss": 0.7753,
+      "step": 141
+    },
+    {
+      "epoch": 0.44842105263157894,
+      "grad_norm": 0.818867027759552,
+      "learning_rate": 9.925277273640211e-06,
+      "loss": 0.7884,
+      "step": 142
+    },
+    {
+      "epoch": 0.4515789473684211,
+      "grad_norm": 0.9297605156898499,
+      "learning_rate": 9.922072129676644e-06,
+      "loss": 0.8113,
+      "step": 143
+    },
+    {
+      "epoch": 0.45473684210526316,
+      "grad_norm": 0.9068427681922913,
+      "learning_rate": 9.918800220665035e-06,
+      "loss": 0.808,
+      "step": 144
+    },
+    {
+      "epoch": 0.45789473684210524,
+      "grad_norm": 0.7996309399604797,
+      "learning_rate": 9.915461590986926e-06,
+      "loss": 0.7967,
+      "step": 145
+    },
+    {
+      "epoch": 0.4610526315789474,
+      "grad_norm": 0.777631938457489,
+      "learning_rate": 9.912056285928891e-06,
+      "loss": 0.8123,
+      "step": 146
+    },
+    {
+      "epoch": 0.46421052631578946,
+      "grad_norm": 0.7640287280082703,
+      "learning_rate": 9.908584351681911e-06,
+      "loss": 0.7942,
+      "step": 147
+    },
+    {
+      "epoch": 0.4673684210526316,
+      "grad_norm": 0.900375247001648,
+      "learning_rate": 9.90504583534076e-06,
+      "loss": 0.7916,
+      "step": 148
+    },
+    {
+      "epoch": 0.4705263157894737,
+      "grad_norm": 0.8545629382133484,
+      "learning_rate": 9.901440784903354e-06,
+      "loss": 0.7946,
+      "step": 149
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.739754319190979,
+      "learning_rate": 9.897769249270106e-06,
+      "loss": 0.797,
+      "step": 150
+    },
+    {
+      "epoch": 0.4768421052631579,
+      "grad_norm": 0.814207136631012,
+      "learning_rate": 9.894031278243266e-06,
+      "loss": 0.7636,
+      "step": 151
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8900481462478638,
+      "learning_rate": 9.890226922526238e-06,
+      "loss": 0.7918,
+      "step": 152
+    },
+    {
+      "epoch": 0.4831578947368421,
+      "grad_norm": 0.7793235182762146,
+      "learning_rate": 9.886356233722894e-06,
+      "loss": 0.7956,
+      "step": 153
+    },
+    {
+      "epoch": 0.4863157894736842,
+      "grad_norm": 0.890493631362915,
+      "learning_rate": 9.88241926433688e-06,
+      "loss": 0.7643,
+      "step": 154
+    },
+    {
+      "epoch": 0.48947368421052634,
+      "grad_norm": 0.8794576525688171,
+      "learning_rate": 9.878416067770898e-06,
+      "loss": 0.7929,
+      "step": 155
+    },
+    {
+      "epoch": 0.4926315789473684,
+      "grad_norm": 0.7571336627006531,
+      "learning_rate": 9.874346698325983e-06,
+      "loss": 0.7852,
+      "step": 156
+    },
+    {
+      "epoch": 0.4957894736842105,
+      "grad_norm": 0.8208326101303101,
+      "learning_rate": 9.870211211200766e-06,
+      "loss": 0.8021,
+      "step": 157
+    },
+    {
+      "epoch": 0.49894736842105264,
+      "grad_norm": 0.904834508895874,
+      "learning_rate": 9.866009662490727e-06,
+      "loss": 0.7809,
+      "step": 158
+    },
+    {
+      "epoch": 0.5021052631578947,
+      "grad_norm": 0.8771081566810608,
+      "learning_rate": 9.861742109187433e-06,
+      "loss": 0.8044,
+      "step": 159
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 0.7162814140319824,
+      "learning_rate": 9.857408609177763e-06,
+      "loss": 0.7935,
+      "step": 160
+    },
+    {
+      "epoch": 0.508421052631579,
+      "grad_norm": 0.8736246824264526,
+      "learning_rate": 9.853009221243129e-06,
+      "loss": 0.781,
+      "step": 161
+    },
+    {
+      "epoch": 0.511578947368421,
+      "grad_norm": 0.7425425052642822,
+      "learning_rate": 9.848544005058668e-06,
+      "loss": 0.7619,
+      "step": 162
+    },
+    {
+      "epoch": 0.5147368421052632,
+      "grad_norm": 0.7853212356567383,
+      "learning_rate": 9.844013021192447e-06,
+      "loss": 0.77,
+      "step": 163
+    },
+    {
+      "epoch": 0.5178947368421053,
+      "grad_norm": 0.7985186576843262,
+      "learning_rate": 9.839416331104625e-06,
+      "loss": 0.7823,
+      "step": 164
+    },
+    {
+      "epoch": 0.5210526315789473,
+      "grad_norm": 0.8637293577194214,
+      "learning_rate": 9.834753997146633e-06,
+      "loss": 0.7598,
+      "step": 165
+    },
+    {
+      "epoch": 0.5242105263157895,
+      "grad_norm": 0.7811316847801208,
+      "learning_rate": 9.830026082560324e-06,
+      "loss": 0.7787,
+      "step": 166
+    },
+    {
+      "epoch": 0.5273684210526316,
+      "grad_norm": 0.795226514339447,
+      "learning_rate": 9.825232651477109e-06,
+      "loss": 0.7843,
+      "step": 167
+    },
+    {
+      "epoch": 0.5305263157894737,
+      "grad_norm": 0.9816672205924988,
+      "learning_rate": 9.820373768917095e-06,
+      "loss": 0.778,
+      "step": 168
+    },
+    {
+      "epoch": 0.5336842105263158,
+      "grad_norm": 0.7347379922866821,
+      "learning_rate": 9.815449500788203e-06,
+      "loss": 0.8002,
+      "step": 169
+    },
+    {
+      "epoch": 0.5368421052631579,
+      "grad_norm": 0.8775848746299744,
+      "learning_rate": 9.810459913885265e-06,
+      "loss": 0.7484,
+      "step": 170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8044248819351196,
+      "learning_rate": 9.805405075889129e-06,
+      "loss": 0.7607,
+      "step": 171
+    },
+    {
+      "epoch": 0.5431578947368421,
+      "grad_norm": 0.8173661231994629,
+      "learning_rate": 9.800285055365737e-06,
+      "loss": 0.7654,
+      "step": 172
+    },
+    {
+      "epoch": 0.5463157894736842,
+      "grad_norm": 0.7608327269554138,
+      "learning_rate": 9.79509992176519e-06,
+      "loss": 0.7716,
+      "step": 173
+    },
+    {
+      "epoch": 0.5494736842105263,
+      "grad_norm": 0.7798534631729126,
+      "learning_rate": 9.789849745420811e-06,
+      "loss": 0.7906,
+      "step": 174
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 0.7842231392860413,
+      "learning_rate": 9.784534597548194e-06,
+      "loss": 0.7752,
+      "step": 175
+    },
+    {
+      "epoch": 0.5557894736842105,
+      "grad_norm": 0.7228990793228149,
+      "learning_rate": 9.779154550244228e-06,
+      "loss": 0.7626,
+      "step": 176
+    },
+    {
+      "epoch": 0.5589473684210526,
+      "grad_norm": 0.8362434506416321,
+      "learning_rate": 9.77370967648613e-06,
+      "loss": 0.7547,
+      "step": 177
+    },
+    {
+      "epoch": 0.5621052631578948,
+      "grad_norm": 0.7536957859992981,
+      "learning_rate": 9.768200050130446e-06,
+      "loss": 0.7658,
+      "step": 178
+    },
+    {
+      "epoch": 0.5652631578947368,
+      "grad_norm": 0.7978958487510681,
+      "learning_rate": 9.76262574591206e-06,
+      "loss": 0.7753,
+      "step": 179
+    },
+    {
+      "epoch": 0.5684210526315789,
+      "grad_norm": 0.7004671692848206,
+      "learning_rate": 9.756986839443166e-06,
+      "loss": 0.8162,
+      "step": 180
+    },
+    {
+      "epoch": 0.5715789473684211,
+      "grad_norm": 0.7977942228317261,
+      "learning_rate": 9.751283407212253e-06,
+      "loss": 0.7756,
+      "step": 181
+    },
+    {
+      "epoch": 0.5747368421052632,
+      "grad_norm": 0.7080273628234863,
+      "learning_rate": 9.745515526583066e-06,
+      "loss": 0.7631,
+      "step": 182
+    },
+    {
+      "epoch": 0.5778947368421052,
+      "grad_norm": 0.8024187088012695,
+      "learning_rate": 9.739683275793554e-06,
+      "loss": 0.7908,
+      "step": 183
+    },
+    {
+      "epoch": 0.5810526315789474,
+      "grad_norm": 0.7169841527938843,
+      "learning_rate": 9.73378673395481e-06,
+      "loss": 0.7602,
+      "step": 184
+    },
+    {
+      "epoch": 0.5842105263157895,
+      "grad_norm": 0.7052045464515686,
+      "learning_rate": 9.727825981049994e-06,
+      "loss": 0.7795,
+      "step": 185
+    },
+    {
+      "epoch": 0.5873684210526315,
+      "grad_norm": 0.7155394554138184,
+      "learning_rate": 9.72180109793326e-06,
+      "loss": 0.7919,
+      "step": 186
+    },
+    {
+      "epoch": 0.5905263157894737,
+      "grad_norm": 0.7038570046424866,
+      "learning_rate": 9.715712166328643e-06,
+      "loss": 0.799,
+      "step": 187
+    },
+    {
+      "epoch": 0.5936842105263158,
+      "grad_norm": 0.6392240524291992,
+      "learning_rate": 9.709559268828963e-06,
+      "loss": 0.7868,
+      "step": 188
+    },
+    {
+      "epoch": 0.5968421052631578,
+      "grad_norm": 0.748368501663208,
+      "learning_rate": 9.703342488894699e-06,
+      "loss": 0.7572,
+      "step": 189
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6144984364509583,
+      "learning_rate": 9.697061910852857e-06,
+      "loss": 0.7579,
+      "step": 190
+    },
+    {
+      "epoch": 0.6031578947368421,
+      "grad_norm": 0.7086066007614136,
+      "learning_rate": 9.690717619895828e-06,
+      "loss": 0.7773,
+      "step": 191
+    },
+    {
+      "epoch": 0.6063157894736843,
+      "grad_norm": 0.7267181277275085,
+      "learning_rate": 9.684309702080234e-06,
+      "loss": 0.8062,
+      "step": 192
+    },
+    {
+      "epoch": 0.6094736842105263,
+      "grad_norm": 0.6966003179550171,
+      "learning_rate": 9.677838244325754e-06,
+      "loss": 0.7542,
+      "step": 193
+    },
+    {
+      "epoch": 0.6126315789473684,
+      "grad_norm": 0.7743439674377441,
+      "learning_rate": 9.671303334413952e-06,
+      "loss": 0.7607,
+      "step": 194
+    },
+    {
+      "epoch": 0.6157894736842106,
+      "grad_norm": 0.698222279548645,
+      "learning_rate": 9.664705060987085e-06,
+      "loss": 0.7181,
+      "step": 195
+    },
+    {
+      "epoch": 0.6189473684210526,
+      "grad_norm": 0.738317608833313,
+      "learning_rate": 9.658043513546898e-06,
+      "loss": 0.8108,
+      "step": 196
+    },
+    {
+      "epoch": 0.6221052631578947,
+      "grad_norm": 0.7287928462028503,
+      "learning_rate": 9.651318782453407e-06,
+      "loss": 0.7593,
+      "step": 197
+    },
+    {
+      "epoch": 0.6252631578947369,
+      "grad_norm": 0.7296643853187561,
+      "learning_rate": 9.644530958923683e-06,
+      "loss": 0.7683,
+      "step": 198
+    },
+    {
+      "epoch": 0.628421052631579,
+      "grad_norm": 0.6896747946739197,
+      "learning_rate": 9.637680135030609e-06,
+      "loss": 0.7851,
+      "step": 199
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.722059428691864,
+      "learning_rate": 9.63076640370163e-06,
+      "loss": 0.7837,
+      "step": 200
+    },
+    {
+      "epoch": 0.6347368421052632,
+      "grad_norm": 0.7174692749977112,
+      "learning_rate": 9.623789858717491e-06,
+      "loss": 0.7959,
+      "step": 201
+    },
+    {
+      "epoch": 0.6378947368421053,
+      "grad_norm": 0.769080400466919,
+      "learning_rate": 9.616750594710972e-06,
+      "loss": 0.7711,
+      "step": 202
+    },
+    {
+      "epoch": 0.6410526315789473,
+      "grad_norm": 0.857413113117218,
+      "learning_rate": 9.6096487071656e-06,
+      "loss": 0.7967,
+      "step": 203
+    },
+    {
+      "epoch": 0.6442105263157895,
+      "grad_norm": 0.6896303296089172,
+      "learning_rate": 9.602484292414348e-06,
+      "loss": 0.7541,
+      "step": 204
+    },
+    {
+      "epoch": 0.6473684210526316,
+      "grad_norm": 0.7579444050788879,
+      "learning_rate": 9.595257447638344e-06,
+      "loss": 0.7721,
+      "step": 205
+    },
+    {
+      "epoch": 0.6505263157894737,
+      "grad_norm": 0.8411123752593994,
+      "learning_rate": 9.587968270865534e-06,
+      "loss": 0.8087,
+      "step": 206
+    },
+    {
+      "epoch": 0.6536842105263158,
+      "grad_norm": 0.769237220287323,
+      "learning_rate": 9.580616860969365e-06,
+      "loss": 0.7511,
+      "step": 207
+    },
+    {
+      "epoch": 0.6568421052631579,
+      "grad_norm": 0.6318850517272949,
+      "learning_rate": 9.573203317667442e-06,
+      "loss": 0.7552,
+      "step": 208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7551825046539307,
+      "learning_rate": 9.56572774152017e-06,
+      "loss": 0.7795,
+      "step": 209
+    },
+    {
+      "epoch": 0.6631578947368421,
+      "grad_norm": 0.8102651834487915,
+      "learning_rate": 9.558190233929396e-06,
+      "loss": 0.7973,
+      "step": 210
+    },
+    {
+      "epoch": 0.6663157894736842,
+      "grad_norm": 0.7229571342468262,
+      "learning_rate": 9.55059089713703e-06,
+      "loss": 0.7906,
+      "step": 211
+    },
+    {
+      "epoch": 0.6694736842105263,
+      "grad_norm": 0.6960732936859131,
+      "learning_rate": 9.54292983422366e-06,
+      "loss": 0.755,
+      "step": 212
+    },
+    {
+      "epoch": 0.6726315789473685,
+      "grad_norm": 0.7564675211906433,
+      "learning_rate": 9.53520714910715e-06,
+      "loss": 0.7993,
+      "step": 213
+    },
+    {
+      "epoch": 0.6757894736842105,
+      "grad_norm": 0.7037273645401001,
+      "learning_rate": 9.527422946541238e-06,
+      "loss": 0.7898,
+      "step": 214
+    },
+    {
+      "epoch": 0.6789473684210526,
+      "grad_norm": 0.735377311706543,
+      "learning_rate": 9.519577332114107e-06,
+      "loss": 0.7599,
+      "step": 215
+    },
+    {
+      "epoch": 0.6821052631578948,
+      "grad_norm": 0.6795523762702942,
+      "learning_rate": 9.511670412246956e-06,
+      "loss": 0.7779,
+      "step": 216
+    },
+    {
+      "epoch": 0.6852631578947368,
+      "grad_norm": 0.6760138273239136,
+      "learning_rate": 9.503702294192563e-06,
+      "loss": 0.7759,
+      "step": 217
+    },
+    {
+      "epoch": 0.6884210526315789,
+      "grad_norm": 0.6858091354370117,
+      "learning_rate": 9.495673086033813e-06,
+      "loss": 0.7508,
+      "step": 218
+    },
+    {
+      "epoch": 0.6915789473684211,
+      "grad_norm": 0.8286890983581543,
+      "learning_rate": 9.487582896682252e-06,
+      "loss": 0.7454,
+      "step": 219
+    },
+    {
+      "epoch": 0.6947368421052632,
+      "grad_norm": 0.8973384499549866,
+      "learning_rate": 9.479431835876596e-06,
+      "loss": 0.7616,
+      "step": 220
+    },
+    {
+      "epoch": 0.6978947368421052,
+      "grad_norm": 0.7554185390472412,
+      "learning_rate": 9.471220014181247e-06,
+      "loss": 0.8202,
+      "step": 221
+    },
+    {
+      "epoch": 0.7010526315789474,
+      "grad_norm": 0.7141987681388855,
+      "learning_rate": 9.462947542984795e-06,
+      "loss": 0.7863,
+      "step": 222
+    },
+    {
+      "epoch": 0.7042105263157895,
+      "grad_norm": 0.9621738195419312,
+      "learning_rate": 9.454614534498506e-06,
+      "loss": 0.7748,
+      "step": 223
+    },
+    {
+      "epoch": 0.7073684210526315,
+      "grad_norm": 0.8891428709030151,
+      "learning_rate": 9.446221101754795e-06,
+      "loss": 0.7568,
+      "step": 224
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 0.8448644876480103,
+      "learning_rate": 9.4377673586057e-06,
+      "loss": 0.7561,
+      "step": 225
+    },
+    {
+      "epoch": 0.7136842105263158,
+      "grad_norm": 0.744163453578949,
+      "learning_rate": 9.429253419721335e-06,
+      "loss": 0.8085,
+      "step": 226
+    },
+    {
+      "epoch": 0.716842105263158,
+      "grad_norm": 0.7832540273666382,
+      "learning_rate": 9.420679400588334e-06,
+      "loss": 0.7592,
+      "step": 227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8211809992790222,
+      "learning_rate": 9.412045417508281e-06,
+      "loss": 0.7546,
+      "step": 228
+    },
+    {
+      "epoch": 0.7231578947368421,
+      "grad_norm": 0.7208123803138733,
+      "learning_rate": 9.40335158759614e-06,
+      "loss": 0.7848,
+      "step": 229
+    },
+    {
+      "epoch": 0.7263157894736842,
+      "grad_norm": 0.770294725894928,
+      "learning_rate": 9.394598028778664e-06,
+      "loss": 0.7693,
+      "step": 230
+    },
+    {
+      "epoch": 0.7294736842105263,
+      "grad_norm": 0.7772082686424255,
+      "learning_rate": 9.385784859792787e-06,
+      "loss": 0.7752,
+      "step": 231
+    },
+    {
+      "epoch": 0.7326315789473684,
+      "grad_norm": 0.7851942181587219,
+      "learning_rate": 9.376912200184029e-06,
+      "loss": 0.7858,
+      "step": 232
+    },
+    {
+      "epoch": 0.7357894736842105,
+      "grad_norm": 0.9354181885719299,
+      "learning_rate": 9.367980170304857e-06,
+      "loss": 0.7474,
+      "step": 233
+    },
+    {
+      "epoch": 0.7389473684210527,
+      "grad_norm": 0.7733820676803589,
+      "learning_rate": 9.35898889131307e-06,
+      "loss": 0.7727,
+      "step": 234
+    },
+    {
+      "epoch": 0.7421052631578947,
+      "grad_norm": 0.746245265007019,
+      "learning_rate": 9.349938485170139e-06,
+      "loss": 0.7459,
+      "step": 235
+    },
+    {
+      "epoch": 0.7452631578947368,
+      "grad_norm": 0.8601150512695312,
+      "learning_rate": 9.340829074639566e-06,
+      "loss": 0.768,
+      "step": 236
+    },
+    {
+      "epoch": 0.748421052631579,
+      "grad_norm": 0.7379437685012817,
+      "learning_rate": 9.331660783285208e-06,
+      "loss": 0.7667,
+      "step": 237
+    },
+    {
+      "epoch": 0.751578947368421,
+      "grad_norm": 0.7364124059677124,
+      "learning_rate": 9.322433735469614e-06,
+      "loss": 0.799,
+      "step": 238
+    },
+    {
+      "epoch": 0.7547368421052632,
+      "grad_norm": 0.7081237435340881,
+      "learning_rate": 9.313148056352321e-06,
+      "loss": 0.7464,
+      "step": 239
+    },
+    {
+      "epoch": 0.7578947368421053,
+      "grad_norm": 0.7287322878837585,
+      "learning_rate": 9.303803871888172e-06,
+      "loss": 0.7915,
+      "step": 240
+    },
+    {
+      "epoch": 0.7610526315789473,
+      "grad_norm": 0.7040133476257324,
+      "learning_rate": 9.2944013088256e-06,
+      "loss": 0.7631,
+      "step": 241
+    },
+    {
+      "epoch": 0.7642105263157895,
+      "grad_norm": 0.7509044408798218,
+      "learning_rate": 9.284940494704906e-06,
+      "loss": 0.7836,
+      "step": 242
+    },
+    {
+      "epoch": 0.7673684210526316,
+      "grad_norm": 0.7864567041397095,
+      "learning_rate": 9.275421557856536e-06,
+      "loss": 0.7547,
+      "step": 243
+    },
+    {
+      "epoch": 0.7705263157894737,
+      "grad_norm": 0.696180522441864,
+      "learning_rate": 9.26584462739934e-06,
+      "loss": 0.7908,
+      "step": 244
+    },
+    {
+      "epoch": 0.7736842105263158,
+      "grad_norm": 0.7591786980628967,
+      "learning_rate": 9.25620983323881e-06,
+      "loss": 0.7663,
+      "step": 245
+    },
+    {
+      "epoch": 0.7768421052631579,
+      "grad_norm": 0.6878987550735474,
+      "learning_rate": 9.246517306065332e-06,
+      "loss": 0.7363,
+      "step": 246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7305520176887512,
+      "learning_rate": 9.236767177352403e-06,
+      "loss": 0.7649,
+      "step": 247
+    },
+    {
+      "epoch": 0.783157894736842,
+      "grad_norm": 0.7379259467124939,
+      "learning_rate": 9.226959579354855e-06,
+      "loss": 0.7733,
+      "step": 248
+    },
+    {
+      "epoch": 0.7863157894736842,
+      "grad_norm": 0.7051509618759155,
+      "learning_rate": 9.217094645107052e-06,
+      "loss": 0.771,
+      "step": 249
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.7959958910942078,
+      "learning_rate": 9.207172508421099e-06,
+      "loss": 0.7529,
+      "step": 250
+    },
+    {
+      "epoch": 0.7926315789473685,
+      "grad_norm": 0.6911975145339966,
+      "learning_rate": 9.197193303885008e-06,
+      "loss": 0.7722,
+      "step": 251
+    },
+    {
+      "epoch": 0.7957894736842105,
+      "grad_norm": 0.8246911764144897,
+      "learning_rate": 9.187157166860894e-06,
+      "loss": 0.768,
+      "step": 252
+    },
+    {
+      "epoch": 0.7989473684210526,
+      "grad_norm": 0.7509826421737671,
+      "learning_rate": 9.177064233483121e-06,
+      "loss": 0.7869,
+      "step": 253
+    },
+    {
+      "epoch": 0.8021052631578948,
+      "grad_norm": 0.7473223209381104,
+      "learning_rate": 9.166914640656467e-06,
+      "loss": 0.7515,
+      "step": 254
+    },
+    {
+      "epoch": 0.8052631578947368,
+      "grad_norm": 0.8410593867301941,
+      "learning_rate": 9.156708526054257e-06,
+      "loss": 0.8106,
+      "step": 255
+    },
+    {
+      "epoch": 0.8084210526315789,
+      "grad_norm": 0.7072636485099792,
+      "learning_rate": 9.146446028116508e-06,
+      "loss": 0.7475,
+      "step": 256
+    },
+    {
+      "epoch": 0.8115789473684211,
+      "grad_norm": 0.7030969858169556,
+      "learning_rate": 9.136127286048038e-06,
+      "loss": 0.7567,
+      "step": 257
+    },
+    {
+      "epoch": 0.8147368421052632,
+      "grad_norm": 0.8154286742210388,
+      "learning_rate": 9.125752439816588e-06,
+      "loss": 0.7884,
+      "step": 258
+    },
+    {
+      "epoch": 0.8178947368421052,
+      "grad_norm": 0.6962621212005615,
+      "learning_rate": 9.115321630150918e-06,
+      "loss": 0.7688,
+      "step": 259
+    },
+    {
+      "epoch": 0.8210526315789474,
+      "grad_norm": 0.7069746255874634,
+      "learning_rate": 9.104834998538899e-06,
+      "loss": 0.7313,
+      "step": 260
+    },
+    {
+      "epoch": 0.8242105263157895,
+      "grad_norm": 0.7636252045631409,
+      "learning_rate": 9.094292687225594e-06,
+      "loss": 0.7617,
+      "step": 261
+    },
+    {
+      "epoch": 0.8273684210526315,
+      "grad_norm": 0.7339772582054138,
+      "learning_rate": 9.08369483921133e-06,
+      "loss": 0.7469,
+      "step": 262
+    },
+    {
+      "epoch": 0.8305263157894737,
+      "grad_norm": 0.71903395652771,
+      "learning_rate": 9.073041598249757e-06,
+      "loss": 0.7662,
+      "step": 263
+    },
+    {
+      "epoch": 0.8336842105263158,
+      "grad_norm": 0.7993423938751221,
+      "learning_rate": 9.062333108845897e-06,
+      "loss": 0.7514,
+      "step": 264
+    },
+    {
+      "epoch": 0.8368421052631579,
+      "grad_norm": 0.8029423356056213,
+      "learning_rate": 9.051569516254186e-06,
+      "loss": 0.7457,
+      "step": 265
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6745309233665466,
+      "learning_rate": 9.040750966476502e-06,
+      "loss": 0.7591,
+      "step": 266
+    },
+    {
+      "epoch": 0.8431578947368421,
+      "grad_norm": 0.8080049753189087,
+      "learning_rate": 9.029877606260187e-06,
+      "loss": 0.7712,
+      "step": 267
+    },
+    {
+      "epoch": 0.8463157894736842,
+      "grad_norm": 0.6188151836395264,
+      "learning_rate": 9.018949583096051e-06,
+      "loss": 0.7339,
+      "step": 268
+    },
+    {
+      "epoch": 0.8494736842105263,
+      "grad_norm": 0.7027499079704285,
+      "learning_rate": 9.00796704521638e-06,
+      "loss": 0.7651,
+      "step": 269
+    },
+    {
+      "epoch": 0.8526315789473684,
+      "grad_norm": 0.6698938608169556,
+      "learning_rate": 8.996930141592915e-06,
+      "loss": 0.7648,
+      "step": 270
+    },
+    {
+      "epoch": 0.8557894736842105,
+      "grad_norm": 0.610205352306366,
+      "learning_rate": 8.985839021934843e-06,
+      "loss": 0.7663,
+      "step": 271
+    },
+    {
+      "epoch": 0.8589473684210527,
+      "grad_norm": 0.6673887372016907,
+      "learning_rate": 8.974693836686755e-06,
+      "loss": 0.7512,
+      "step": 272
+    },
+    {
+      "epoch": 0.8621052631578947,
+      "grad_norm": 0.6559451818466187,
+      "learning_rate": 8.963494737026612e-06,
+      "loss": 0.7553,
+      "step": 273
+    },
+    {
+      "epoch": 0.8652631578947368,
+      "grad_norm": 0.679578959941864,
+      "learning_rate": 8.952241874863695e-06,
+      "loss": 0.7456,
+      "step": 274
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 0.6281573176383972,
+      "learning_rate": 8.940935402836535e-06,
+      "loss": 0.7251,
+      "step": 275
+    },
+    {
+      "epoch": 0.871578947368421,
+      "grad_norm": 0.6603413820266724,
+      "learning_rate": 8.92957547431086e-06,
+      "loss": 0.7518,
+      "step": 276
+    },
+    {
+      "epoch": 0.8747368421052631,
+      "grad_norm": 0.759972870349884,
+      "learning_rate": 8.918162243377494e-06,
+      "loss": 0.755,
+      "step": 277
+    },
+    {
+      "epoch": 0.8778947368421053,
+      "grad_norm": 0.7080827951431274,
+      "learning_rate": 8.906695864850284e-06,
+      "loss": 0.7868,
+      "step": 278
+    },
+    {
+      "epoch": 0.8810526315789474,
+      "grad_norm": 0.6595768928527832,
+      "learning_rate": 8.895176494263993e-06,
+      "loss": 0.7536,
+      "step": 279
+    },
+    {
+      "epoch": 0.8842105263157894,
+      "grad_norm": 0.6742393374443054,
+      "learning_rate": 8.883604287872186e-06,
+      "loss": 0.7771,
+      "step": 280
+    },
+    {
+      "epoch": 0.8873684210526316,
+      "grad_norm": 0.7637141942977905,
+      "learning_rate": 8.871979402645116e-06,
+      "loss": 0.7723,
+      "step": 281
+    },
+    {
+      "epoch": 0.8905263157894737,
+      "grad_norm": 0.6951268315315247,
+      "learning_rate": 8.860301996267601e-06,
+      "loss": 0.7668,
+      "step": 282
+    },
+    {
+      "epoch": 0.8936842105263157,
+      "grad_norm": 0.672383189201355,
+      "learning_rate": 8.848572227136869e-06,
+      "loss": 0.759,
+      "step": 283
+    },
+    {
+      "epoch": 0.8968421052631579,
+      "grad_norm": 0.8001788854598999,
+      "learning_rate": 8.83679025436042e-06,
+      "loss": 0.7903,
+      "step": 284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7648778557777405,
+      "learning_rate": 8.824956237753872e-06,
+      "loss": 0.7746,
+      "step": 285
+    },
+    {
+      "epoch": 0.9031578947368422,
+      "grad_norm": 0.6406739354133606,
+      "learning_rate": 8.813070337838781e-06,
+      "loss": 0.7771,
+      "step": 286
+    },
+    {
+      "epoch": 0.9063157894736842,
+      "grad_norm": 0.6822999715805054,
+      "learning_rate": 8.80113271584047e-06,
+      "loss": 0.7592,
+      "step": 287
+    },
+    {
+      "epoch": 0.9094736842105263,
+      "grad_norm": 0.7665005326271057,
+      "learning_rate": 8.789143533685847e-06,
+      "loss": 0.747,
+      "step": 288
+    },
+    {
+      "epoch": 0.9126315789473685,
+      "grad_norm": 0.7119230628013611,
+      "learning_rate": 8.777102954001199e-06,
+      "loss": 0.7686,
+      "step": 289
+    },
+    {
+      "epoch": 0.9157894736842105,
+      "grad_norm": 0.6693398356437683,
+      "learning_rate": 8.765011140109993e-06,
+      "loss": 0.7496,
+      "step": 290
+    },
+    {
+      "epoch": 0.9189473684210526,
+      "grad_norm": 0.6795002818107605,
+      "learning_rate": 8.752868256030658e-06,
+      "loss": 0.7543,
+      "step": 291
+    },
+    {
+      "epoch": 0.9221052631578948,
+      "grad_norm": 0.773586094379425,
+      "learning_rate": 8.740674466474357e-06,
+      "loss": 0.7822,
+      "step": 292
+    },
+    {
+      "epoch": 0.9252631578947368,
+      "grad_norm": 0.7866982817649841,
+      "learning_rate": 8.728429936842762e-06,
+      "loss": 0.7792,
+      "step": 293
+    },
+    {
+      "epoch": 0.9284210526315789,
+      "grad_norm": 0.6205121874809265,
+      "learning_rate": 8.716134833225803e-06,
+      "loss": 0.7541,
+      "step": 294
+    },
+    {
+      "epoch": 0.9315789473684211,
+      "grad_norm": 0.819891095161438,
+      "learning_rate": 8.70378932239941e-06,
+      "loss": 0.7477,
+      "step": 295
+    },
+    {
+      "epoch": 0.9347368421052632,
+      "grad_norm": 0.7897220253944397,
+      "learning_rate": 8.691393571823266e-06,
+      "loss": 0.7401,
+      "step": 296
+    },
+    {
+      "epoch": 0.9378947368421052,
+      "grad_norm": 0.772665798664093,
+      "learning_rate": 8.678947749638525e-06,
+      "loss": 0.7888,
+      "step": 297
+    },
+    {
+      "epoch": 0.9410526315789474,
+      "grad_norm": 0.7343323826789856,
+      "learning_rate": 8.666452024665533e-06,
+      "loss": 0.7321,
+      "step": 298
+    },
+    {
+      "epoch": 0.9442105263157895,
+      "grad_norm": 0.9545820355415344,
+      "learning_rate": 8.653906566401533e-06,
+      "loss": 0.761,
+      "step": 299
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.8397694230079651,
+      "learning_rate": 8.64131154501838e-06,
+      "loss": 0.7766,
+      "step": 300
+    },
+    {
+      "epoch": 0.9505263157894737,
+      "grad_norm": 0.7189294695854187,
+      "learning_rate": 8.628667131360218e-06,
+      "loss": 0.7279,
+      "step": 301
+    },
+    {
+      "epoch": 0.9536842105263158,
+      "grad_norm": 0.7687074542045593,
+      "learning_rate": 8.61597349694117e-06,
+      "loss": 0.8155,
+      "step": 302
+    },
+    {
+      "epoch": 0.9568421052631579,
+      "grad_norm": 0.7857553362846375,
+      "learning_rate": 8.60323081394301e-06,
+      "loss": 0.7422,
+      "step": 303
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7681856751441956,
+      "learning_rate": 8.59043925521283e-06,
+      "loss": 0.7551,
+      "step": 304
+    },
+    {
+      "epoch": 0.9631578947368421,
+      "grad_norm": 0.7847979068756104,
+      "learning_rate": 8.577598994260687e-06,
+      "loss": 0.7978,
+      "step": 305
+    },
+    {
+      "epoch": 0.9663157894736842,
+      "grad_norm": 0.7032335996627808,
+      "learning_rate": 8.56471020525726e-06,
+      "loss": 0.7596,
+      "step": 306
+    },
+    {
+      "epoch": 0.9694736842105263,
+      "grad_norm": 0.8031393885612488,
+      "learning_rate": 8.551773063031484e-06,
+      "loss": 0.7615,
+      "step": 307
+    },
+    {
+      "epoch": 0.9726315789473684,
+      "grad_norm": 0.9480533599853516,
+      "learning_rate": 8.538787743068172e-06,
+      "loss": 0.7728,
+      "step": 308
+    },
+    {
+      "epoch": 0.9757894736842105,
+      "grad_norm": 0.7876819968223572,
+      "learning_rate": 8.525754421505646e-06,
+      "loss": 0.7871,
+      "step": 309
+    },
+    {
+      "epoch": 0.9789473684210527,
+      "grad_norm": 0.8743879199028015,
+      "learning_rate": 8.512673275133334e-06,
+      "loss": 0.7953,
+      "step": 310
+    },
+    {
+      "epoch": 0.9821052631578947,
+      "grad_norm": 0.8548572063446045,
+      "learning_rate": 8.49954448138939e-06,
+      "loss": 0.751,
+      "step": 311
+    },
+    {
+      "epoch": 0.9852631578947368,
+      "grad_norm": 0.801235020160675,
+      "learning_rate": 8.486368218358268e-06,
+      "loss": 0.7325,
+      "step": 312
+    },
+    {
+      "epoch": 0.988421052631579,
+      "grad_norm": 0.6868360042572021,
+      "learning_rate": 8.473144664768322e-06,
+      "loss": 0.7565,
+      "step": 313
+    },
+    {
+      "epoch": 0.991578947368421,
+      "grad_norm": 0.6946709156036377,
+      "learning_rate": 8.459873999989367e-06,
+      "loss": 0.7436,
+      "step": 314
+    },
+    {
+      "epoch": 0.9947368421052631,
+      "grad_norm": 0.7966560125350952,
+      "learning_rate": 8.446556404030263e-06,
+      "loss": 0.7818,
+      "step": 315
+    },
+    {
+      "epoch": 0.9978947368421053,
+      "grad_norm": 0.7011910080909729,
+      "learning_rate": 8.433192057536458e-06,
+      "loss": 0.8047,
+      "step": 316
+    },
+    {
+      "epoch": 1.0010526315789474,
+      "grad_norm": 0.776292085647583,
+      "learning_rate": 8.419781141787549e-06,
+      "loss": 0.7393,
+      "step": 317
+    },
+    {
+      "epoch": 1.0042105263157894,
+      "grad_norm": 0.7284736037254333,
+      "learning_rate": 8.406323838694808e-06,
+      "loss": 0.6665,
+      "step": 318
+    },
+    {
+      "epoch": 1.0073684210526315,
+      "grad_norm": 0.6233306527137756,
+      "learning_rate": 8.392820330798734e-06,
+      "loss": 0.6677,
+      "step": 319
+    },
+    {
+      "epoch": 1.0105263157894737,
+      "grad_norm": 0.6907612085342407,
+      "learning_rate": 8.379270801266569e-06,
+      "loss": 0.6921,
+      "step": 320
+    },
+    {
+      "epoch": 1.0136842105263157,
+      "grad_norm": 0.7600920796394348,
+      "learning_rate": 8.3656754338898e-06,
+      "loss": 0.7129,
+      "step": 321
+    },
+    {
+      "epoch": 1.016842105263158,
+      "grad_norm": 0.7222225069999695,
+      "learning_rate": 8.352034413081687e-06,
+      "loss": 0.6633,
+      "step": 322
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8816447854042053,
+      "learning_rate": 8.33834792387475e-06,
+      "loss": 0.7188,
+      "step": 323
+    },
+    {
+      "epoch": 1.023157894736842,
+      "grad_norm": 0.7416877150535583,
+      "learning_rate": 8.324616151918263e-06,
+      "loss": 0.7279,
+      "step": 324
+    },
+    {
+      "epoch": 1.0263157894736843,
+      "grad_norm": 0.7744479775428772,
+      "learning_rate": 8.31083928347573e-06,
+      "loss": 0.6765,
+      "step": 325
+    },
+    {
+      "epoch": 1.0294736842105263,
+      "grad_norm": 0.7218484282493591,
+      "learning_rate": 8.297017505422366e-06,
+      "loss": 0.6872,
+      "step": 326
+    },
+    {
+      "epoch": 1.0326315789473683,
+      "grad_norm": 0.6902859807014465,
+      "learning_rate": 8.28315100524256e-06,
+      "loss": 0.6696,
+      "step": 327
+    },
+    {
+      "epoch": 1.0357894736842106,
+      "grad_norm": 0.6810570359230042,
+      "learning_rate": 8.269239971027328e-06,
+      "loss": 0.692,
+      "step": 328
+    },
+    {
+      "epoch": 1.0389473684210526,
+      "grad_norm": 0.7604600191116333,
+      "learning_rate": 8.255284591471762e-06,
+      "loss": 0.7042,
+      "step": 329
+    },
+    {
+      "epoch": 1.0421052631578946,
+      "grad_norm": 0.6768532991409302,
+      "learning_rate": 8.241285055872478e-06,
+      "loss": 0.7111,
+      "step": 330
+    },
+    {
+      "epoch": 1.045263157894737,
+      "grad_norm": 0.7930629849433899,
+      "learning_rate": 8.227241554125041e-06,
+      "loss": 0.7196,
+      "step": 331
+    },
+    {
+      "epoch": 1.048421052631579,
+      "grad_norm": 0.7775673270225525,
+      "learning_rate": 8.213154276721388e-06,
+      "loss": 0.6706,
+      "step": 332
+    },
+    {
+      "epoch": 1.051578947368421,
+      "grad_norm": 0.7330057621002197,
+      "learning_rate": 8.199023414747257e-06,
+      "loss": 0.6841,
+      "step": 333
+    },
+    {
+      "epoch": 1.0547368421052632,
+      "grad_norm": 0.6644675731658936,
+      "learning_rate": 8.18484915987957e-06,
+      "loss": 0.6813,
+      "step": 334
+    },
+    {
+      "epoch": 1.0578947368421052,
+      "grad_norm": 0.7848517894744873,
+      "learning_rate": 8.170631704383865e-06,
+      "loss": 0.6907,
+      "step": 335
+    },
+    {
+      "epoch": 1.0610526315789475,
+      "grad_norm": 0.6873379349708557,
+      "learning_rate": 8.15637124111166e-06,
+      "loss": 0.7008,
+      "step": 336
+    },
+    {
+      "epoch": 1.0642105263157895,
+      "grad_norm": 0.6563950777053833,
+      "learning_rate": 8.14206796349785e-06,
+      "loss": 0.7082,
+      "step": 337
+    },
+    {
+      "epoch": 1.0673684210526315,
+      "grad_norm": 0.6331363916397095,
+      "learning_rate": 8.127722065558087e-06,
+      "loss": 0.6702,
+      "step": 338
+    },
+    {
+      "epoch": 1.0705263157894738,
+      "grad_norm": 0.7190977931022644,
+      "learning_rate": 8.113333741886137e-06,
+      "loss": 0.6687,
+      "step": 339
+    },
+    {
+      "epoch": 1.0736842105263158,
+      "grad_norm": 0.6932594180107117,
+      "learning_rate": 8.098903187651252e-06,
+      "loss": 0.6677,
+      "step": 340
+    },
+    {
+      "epoch": 1.0768421052631578,
+      "grad_norm": 0.7676414251327515,
+      "learning_rate": 8.084430598595514e-06,
+      "loss": 0.6945,
+      "step": 341
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7621206045150757,
+      "learning_rate": 8.069916171031181e-06,
+      "loss": 0.6908,
+      "step": 342
+    },
+    {
+      "epoch": 1.083157894736842,
+      "grad_norm": 0.7197576761245728,
+      "learning_rate": 8.055360101838026e-06,
+      "loss": 0.7144,
+      "step": 343
+    },
+    {
+      "epoch": 1.0863157894736841,
+      "grad_norm": 0.7100708484649658,
+      "learning_rate": 8.04076258846067e-06,
+      "loss": 0.6983,
+      "step": 344
+    },
+    {
+      "epoch": 1.0894736842105264,
+      "grad_norm": 0.6768266558647156,
+      "learning_rate": 8.026123828905902e-06,
+      "loss": 0.6769,
+      "step": 345
+    },
+    {
+      "epoch": 1.0926315789473684,
+      "grad_norm": 0.7214046716690063,
+      "learning_rate": 8.011444021739986e-06,
+      "loss": 0.7022,
+      "step": 346
+    },
+    {
+      "epoch": 1.0957894736842104,
+      "grad_norm": 0.7675743699073792,
+      "learning_rate": 7.996723366085978e-06,
+      "loss": 0.6718,
+      "step": 347
+    },
+    {
+      "epoch": 1.0989473684210527,
+      "grad_norm": 0.7916297912597656,
+      "learning_rate": 7.981962061621012e-06,
+      "loss": 0.6947,
+      "step": 348
+    },
+    {
+      "epoch": 1.1021052631578947,
+      "grad_norm": 0.7070953845977783,
+      "learning_rate": 7.967160308573607e-06,
+      "loss": 0.6223,
+      "step": 349
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.8087365031242371,
+      "learning_rate": 7.952318307720943e-06,
+      "loss": 0.724,
+      "step": 350
+    },
+    {
+      "epoch": 1.108421052631579,
+      "grad_norm": 0.6387585401535034,
+      "learning_rate": 7.937436260386134e-06,
+      "loss": 0.6732,
+      "step": 351
+    },
+    {
+      "epoch": 1.111578947368421,
+      "grad_norm": 0.7180246114730835,
+      "learning_rate": 7.922514368435506e-06,
+      "loss": 0.669,
+      "step": 352
+    },
+    {
+      "epoch": 1.1147368421052632,
+      "grad_norm": 0.7500141859054565,
+      "learning_rate": 7.907552834275847e-06,
+      "loss": 0.6946,
+      "step": 353
+    },
+    {
+      "epoch": 1.1178947368421053,
+      "grad_norm": 0.6470015048980713,
+      "learning_rate": 7.892551860851679e-06,
+      "loss": 0.6826,
+      "step": 354
+    },
+    {
+      "epoch": 1.1210526315789473,
+      "grad_norm": 0.668079674243927,
+      "learning_rate": 7.877511651642486e-06,
+      "loss": 0.6672,
+      "step": 355
+    },
+    {
+      "epoch": 1.1242105263157895,
+      "grad_norm": 0.7336689829826355,
+      "learning_rate": 7.862432410659964e-06,
+      "loss": 0.6932,
+      "step": 356
+    },
+    {
+      "epoch": 1.1273684210526316,
+      "grad_norm": 0.6308228969573975,
+      "learning_rate": 7.847314342445258e-06,
+      "loss": 0.7253,
+      "step": 357
+    },
+    {
+      "epoch": 1.1305263157894736,
+      "grad_norm": 0.6617457866668701,
+      "learning_rate": 7.832157652066173e-06,
+      "loss": 0.7053,
+      "step": 358
+    },
+    {
+      "epoch": 1.1336842105263158,
+      "grad_norm": 0.7797347903251648,
+      "learning_rate": 7.81696254511441e-06,
+      "loss": 0.6927,
+      "step": 359
+    },
+    {
+      "epoch": 1.1368421052631579,
+      "grad_norm": 0.6679685711860657,
+      "learning_rate": 7.80172922770276e-06,
+      "loss": 0.6917,
+      "step": 360
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 0.7145143151283264,
+      "learning_rate": 7.786457906462329e-06,
+      "loss": 0.6967,
+      "step": 361
+    },
+    {
+      "epoch": 1.1431578947368422,
+      "grad_norm": 0.8625593185424805,
+      "learning_rate": 7.771148788539704e-06,
+      "loss": 0.6728,
+      "step": 362
+    },
+    {
+      "epoch": 1.1463157894736842,
+      "grad_norm": 0.7024162411689758,
+      "learning_rate": 7.755802081594179e-06,
+      "loss": 0.7053,
+      "step": 363
+    },
+    {
+      "epoch": 1.1494736842105264,
+      "grad_norm": 0.7329344153404236,
+      "learning_rate": 7.740417993794918e-06,
+      "loss": 0.644,
+      "step": 364
+    },
+    {
+      "epoch": 1.1526315789473685,
+      "grad_norm": 0.7487715482711792,
+      "learning_rate": 7.724996733818124e-06,
+      "loss": 0.6838,
+      "step": 365
+    },
+    {
+      "epoch": 1.1557894736842105,
+      "grad_norm": 0.7509961128234863,
+      "learning_rate": 7.709538510844234e-06,
+      "loss": 0.6527,
+      "step": 366
+    },
+    {
+      "epoch": 1.1589473684210527,
+      "grad_norm": 0.7864775657653809,
+      "learning_rate": 7.694043534555055e-06,
+      "loss": 0.7205,
+      "step": 367
+    },
+    {
+      "epoch": 1.1621052631578948,
+      "grad_norm": 0.821261465549469,
+      "learning_rate": 7.678512015130936e-06,
+      "loss": 0.6832,
+      "step": 368
+    },
+    {
+      "epoch": 1.1652631578947368,
+      "grad_norm": 0.7063290476799011,
+      "learning_rate": 7.662944163247916e-06,
+      "loss": 0.6964,
+      "step": 369
+    },
+    {
+      "epoch": 1.168421052631579,
+      "grad_norm": 0.7851418256759644,
+      "learning_rate": 7.647340190074854e-06,
+      "loss": 0.7028,
+      "step": 370
+    },
+    {
+      "epoch": 1.171578947368421,
+      "grad_norm": 0.747036337852478,
+      "learning_rate": 7.63170030727058e-06,
+      "loss": 0.637,
+      "step": 371
+    },
+    {
+      "epoch": 1.174736842105263,
+      "grad_norm": 0.6792373657226562,
+      "learning_rate": 7.616024726981015e-06,
+      "loss": 0.6903,
+      "step": 372
+    },
+    {
+      "epoch": 1.1778947368421053,
+      "grad_norm": 0.805324912071228,
+      "learning_rate": 7.600313661836298e-06,
+      "loss": 0.6826,
+      "step": 373
+    },
+    {
+      "epoch": 1.1810526315789474,
+      "grad_norm": 0.6382725238800049,
+      "learning_rate": 7.584567324947893e-06,
+      "loss": 0.6337,
+      "step": 374
+    },
+    {
+      "epoch": 1.1842105263157894,
+      "grad_norm": 0.606036365032196,
+      "learning_rate": 7.568785929905713e-06,
+      "loss": 0.7219,
+      "step": 375
+    },
+    {
+      "epoch": 1.1873684210526316,
+      "grad_norm": 0.6815652847290039,
+      "learning_rate": 7.552969690775209e-06,
+      "loss": 0.6729,
+      "step": 376
+    },
+    {
+      "epoch": 1.1905263157894737,
+      "grad_norm": 0.6756043434143066,
+      "learning_rate": 7.537118822094474e-06,
+      "loss": 0.7,
+      "step": 377
+    },
+    {
+      "epoch": 1.1936842105263157,
+      "grad_norm": 0.6447840929031372,
+      "learning_rate": 7.521233538871329e-06,
+      "loss": 0.6982,
+      "step": 378
+    },
+    {
+      "epoch": 1.196842105263158,
+      "grad_norm": 0.5919721126556396,
+      "learning_rate": 7.505314056580411e-06,
+      "loss": 0.6923,
+      "step": 379
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6429916620254517,
+      "learning_rate": 7.489360591160245e-06,
+      "loss": 0.6523,
+      "step": 380
+    },
+    {
+      "epoch": 1.203157894736842,
+      "grad_norm": 0.7389147281646729,
+      "learning_rate": 7.4733733590103185e-06,
+      "loss": 0.6774,
+      "step": 381
+    },
+    {
+      "epoch": 1.2063157894736842,
+      "grad_norm": 0.6661043763160706,
+      "learning_rate": 7.457352576988144e-06,
+      "loss": 0.7086,
+      "step": 382
+    },
+    {
+      "epoch": 1.2094736842105263,
+      "grad_norm": 0.6406580209732056,
+      "learning_rate": 7.441298462406321e-06,
+      "loss": 0.7328,
+      "step": 383
+    },
+    {
+      "epoch": 1.2126315789473685,
+      "grad_norm": 0.613018810749054,
+      "learning_rate": 7.4252112330295835e-06,
+      "loss": 0.7078,
+      "step": 384
+    },
+    {
+      "epoch": 1.2157894736842105,
+      "grad_norm": 0.696648359298706,
+      "learning_rate": 7.409091107071849e-06,
+      "loss": 0.7006,
+      "step": 385
+    },
+    {
+      "epoch": 1.2189473684210526,
+      "grad_norm": 0.6505605578422546,
+      "learning_rate": 7.392938303193257e-06,
+      "loss": 0.6805,
+      "step": 386
+    },
+    {
+      "epoch": 1.2221052631578948,
+      "grad_norm": 0.6389300227165222,
+      "learning_rate": 7.376753040497207e-06,
+      "loss": 0.6447,
+      "step": 387
+    },
+    {
+      "epoch": 1.2252631578947368,
+      "grad_norm": 0.6349870562553406,
+      "learning_rate": 7.3605355385273805e-06,
+      "loss": 0.6565,
+      "step": 388
+    },
+    {
+      "epoch": 1.2284210526315789,
+      "grad_norm": 0.6041010618209839,
+      "learning_rate": 7.344286017264765e-06,
+      "loss": 0.708,
+      "step": 389
+    },
+    {
+      "epoch": 1.231578947368421,
+      "grad_norm": 0.6843451261520386,
+      "learning_rate": 7.3280046971246786e-06,
+      "loss": 0.7028,
+      "step": 390
+    },
+    {
+      "epoch": 1.2347368421052631,
+      "grad_norm": 0.6678407788276672,
+      "learning_rate": 7.311691798953765e-06,
+      "loss": 0.702,
+      "step": 391
+    },
+    {
+      "epoch": 1.2378947368421054,
+      "grad_norm": 0.6551707983016968,
+      "learning_rate": 7.295347544027006e-06,
+      "loss": 0.6968,
+      "step": 392
+    },
+    {
+      "epoch": 1.2410526315789474,
+      "grad_norm": 0.6763453483581543,
+      "learning_rate": 7.278972154044722e-06,
+      "loss": 0.7002,
+      "step": 393
+    },
+    {
+      "epoch": 1.2442105263157894,
+      "grad_norm": 0.6329587697982788,
+      "learning_rate": 7.2625658511295635e-06,
+      "loss": 0.6731,
+      "step": 394
+    },
+    {
+      "epoch": 1.2473684210526317,
+      "grad_norm": 0.6213243007659912,
+      "learning_rate": 7.2461288578234955e-06,
+      "loss": 0.6926,
+      "step": 395
+    },
+    {
+      "epoch": 1.2505263157894737,
+      "grad_norm": 0.6473403573036194,
+      "learning_rate": 7.229661397084775e-06,
+      "loss": 0.684,
+      "step": 396
+    },
+    {
+      "epoch": 1.2536842105263157,
+      "grad_norm": 0.6111834049224854,
+      "learning_rate": 7.213163692284943e-06,
+      "loss": 0.7092,
+      "step": 397
+    },
+    {
+      "epoch": 1.256842105263158,
+      "grad_norm": 0.693008303642273,
+      "learning_rate": 7.196635967205776e-06,
+      "loss": 0.7175,
+      "step": 398
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6312058568000793,
+      "learning_rate": 7.180078446036259e-06,
+      "loss": 0.6499,
+      "step": 399
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.6444021463394165,
+      "learning_rate": 7.163491353369545e-06,
+      "loss": 0.6972,
+      "step": 400
+    },
+    {
+      "epoch": 1.2663157894736843,
+      "grad_norm": 0.6179962158203125,
+      "learning_rate": 7.146874914199906e-06,
+      "loss": 0.7037,
+      "step": 401
+    },
+    {
+      "epoch": 1.2694736842105263,
+      "grad_norm": 0.6444630026817322,
+      "learning_rate": 7.130229353919685e-06,
+      "loss": 0.6928,
+      "step": 402
+    },
+    {
+      "epoch": 1.2726315789473683,
+      "grad_norm": 0.666993260383606,
+      "learning_rate": 7.113554898316231e-06,
+      "loss": 0.6872,
+      "step": 403
+    },
+    {
+      "epoch": 1.2757894736842106,
+      "grad_norm": 0.6936898231506348,
+      "learning_rate": 7.0968517735688445e-06,
+      "loss": 0.6788,
+      "step": 404
+    },
+    {
+      "epoch": 1.2789473684210526,
+      "grad_norm": 0.5810768604278564,
+      "learning_rate": 7.080120206245709e-06,
+      "loss": 0.7012,
+      "step": 405
+    },
+    {
+      "epoch": 1.2821052631578946,
+      "grad_norm": 0.6620011925697327,
+      "learning_rate": 7.063360423300808e-06,
+      "loss": 0.6811,
+      "step": 406
+    },
+    {
+      "epoch": 1.2852631578947369,
+      "grad_norm": 0.6566019058227539,
+      "learning_rate": 7.04657265207086e-06,
+      "loss": 0.6951,
+      "step": 407
+    },
+    {
+      "epoch": 1.288421052631579,
+      "grad_norm": 0.6208463311195374,
+      "learning_rate": 7.029757120272222e-06,
+      "loss": 0.7054,
+      "step": 408
+    },
+    {
+      "epoch": 1.291578947368421,
+      "grad_norm": 0.6561064124107361,
+      "learning_rate": 7.0129140559978184e-06,
+      "loss": 0.6849,
+      "step": 409
+    },
+    {
+      "epoch": 1.2947368421052632,
+      "grad_norm": 0.6156150698661804,
+      "learning_rate": 6.99604368771402e-06,
+      "loss": 0.6664,
+      "step": 410
+    },
+    {
+      "epoch": 1.2978947368421052,
+      "grad_norm": 0.7128709554672241,
+      "learning_rate": 6.979146244257573e-06,
+      "loss": 0.6856,
+      "step": 411
+    },
+    {
+      "epoch": 1.3010526315789472,
+      "grad_norm": 0.7587366700172424,
+      "learning_rate": 6.962221954832476e-06,
+      "loss": 0.6407,
+      "step": 412
+    },
+    {
+      "epoch": 1.3042105263157895,
+      "grad_norm": 0.7253438234329224,
+      "learning_rate": 6.945271049006882e-06,
+      "loss": 0.6383,
+      "step": 413
+    },
+    {
+      "epoch": 1.3073684210526315,
+      "grad_norm": 0.7263516187667847,
+      "learning_rate": 6.928293756709976e-06,
+      "loss": 0.7091,
+      "step": 414
+    },
+    {
+      "epoch": 1.3105263157894738,
+      "grad_norm": 0.7259902358055115,
+      "learning_rate": 6.911290308228861e-06,
+      "loss": 0.658,
+      "step": 415
+    },
+    {
+      "epoch": 1.3136842105263158,
+      "grad_norm": 0.6816960573196411,
+      "learning_rate": 6.894260934205437e-06,
+      "loss": 0.6276,
+      "step": 416
+    },
+    {
+      "epoch": 1.3168421052631578,
+      "grad_norm": 0.6776388883590698,
+      "learning_rate": 6.8772058656332626e-06,
+      "loss": 0.7156,
+      "step": 417
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.658682644367218,
+      "learning_rate": 6.860125333854437e-06,
+      "loss": 0.6832,
+      "step": 418
+    },
+    {
+      "epoch": 1.323157894736842,
+      "grad_norm": 0.6834143400192261,
+      "learning_rate": 6.843019570556443e-06,
+      "loss": 0.7103,
+      "step": 419
+    },
+    {
+      "epoch": 1.3263157894736843,
+      "grad_norm": 0.7089201211929321,
+      "learning_rate": 6.82588880776902e-06,
+      "loss": 0.7263,
+      "step": 420
+    },
+    {
+      "epoch": 1.3294736842105264,
+      "grad_norm": 0.6268842220306396,
+      "learning_rate": 6.8087332778610116e-06,
+      "loss": 0.6906,
+      "step": 421
+    },
+    {
+      "epoch": 1.3326315789473684,
+      "grad_norm": 0.6012834906578064,
+      "learning_rate": 6.791553213537209e-06,
+      "loss": 0.7038,
+      "step": 422
+    },
+    {
+      "epoch": 1.3357894736842106,
+      "grad_norm": 0.6678085923194885,
+      "learning_rate": 6.774348847835203e-06,
+      "loss": 0.7062,
+      "step": 423
+    },
+    {
+      "epoch": 1.3389473684210527,
+      "grad_norm": 0.7389436364173889,
+      "learning_rate": 6.757120414122214e-06,
+      "loss": 0.6959,
+      "step": 424
+    },
+    {
+      "epoch": 1.3421052631578947,
+      "grad_norm": 0.6289021968841553,
+      "learning_rate": 6.739868146091934e-06,
+      "loss": 0.7158,
+      "step": 425
+    },
+    {
+      "epoch": 1.345263157894737,
+      "grad_norm": 0.6698178052902222,
+      "learning_rate": 6.722592277761355e-06,
+      "loss": 0.6744,
+      "step": 426
+    },
+    {
+      "epoch": 1.348421052631579,
+      "grad_norm": 0.7639443874359131,
+      "learning_rate": 6.705293043467589e-06,
+      "loss": 0.6726,
+      "step": 427
+    },
+    {
+      "epoch": 1.351578947368421,
+      "grad_norm": 0.6818212270736694,
+      "learning_rate": 6.687970677864696e-06,
+      "loss": 0.7182,
+      "step": 428
+    },
+    {
+      "epoch": 1.3547368421052632,
+      "grad_norm": 0.6343548893928528,
+      "learning_rate": 6.6706254159205e-06,
+      "loss": 0.6823,
+      "step": 429
+    },
+    {
+      "epoch": 1.3578947368421053,
+      "grad_norm": 0.6729276776313782,
+      "learning_rate": 6.653257492913398e-06,
+      "loss": 0.6729,
+      "step": 430
+    },
+    {
+      "epoch": 1.3610526315789473,
+      "grad_norm": 0.6306588053703308,
+      "learning_rate": 6.6358671444291735e-06,
+      "loss": 0.6753,
+      "step": 431
+    },
+    {
+      "epoch": 1.3642105263157895,
+      "grad_norm": 0.5845309495925903,
+      "learning_rate": 6.618454606357796e-06,
+      "loss": 0.6878,
+      "step": 432
+    },
+    {
+      "epoch": 1.3673684210526316,
+      "grad_norm": 0.7026935815811157,
+      "learning_rate": 6.601020114890227e-06,
+      "loss": 0.691,
+      "step": 433
+    },
+    {
+      "epoch": 1.3705263157894736,
+      "grad_norm": 0.7399420738220215,
+      "learning_rate": 6.5835639065152104e-06,
+      "loss": 0.7315,
+      "step": 434
+    },
+    {
+      "epoch": 1.3736842105263158,
+      "grad_norm": 0.6496444344520569,
+      "learning_rate": 6.56608621801607e-06,
+      "loss": 0.7043,
+      "step": 435
+    },
+    {
+      "epoch": 1.3768421052631579,
+      "grad_norm": 0.6120364665985107,
+      "learning_rate": 6.548587286467491e-06,
+      "loss": 0.6685,
+      "step": 436
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6417269706726074,
+      "learning_rate": 6.531067349232314e-06,
+      "loss": 0.6831,
+      "step": 437
+    },
+    {
+      "epoch": 1.3831578947368421,
+      "grad_norm": 0.8120979070663452,
+      "learning_rate": 6.5135266439583015e-06,
+      "loss": 0.6841,
+      "step": 438
+    },
+    {
+      "epoch": 1.3863157894736842,
+      "grad_norm": 0.7373508214950562,
+      "learning_rate": 6.495965408574929e-06,
+      "loss": 0.6976,
+      "step": 439
+    },
+    {
+      "epoch": 1.3894736842105262,
+      "grad_norm": 0.6089206337928772,
+      "learning_rate": 6.478383881290152e-06,
+      "loss": 0.6892,
+      "step": 440
+    },
+    {
+      "epoch": 1.3926315789473684,
+      "grad_norm": 0.7098264098167419,
+      "learning_rate": 6.460782300587166e-06,
+      "loss": 0.7105,
+      "step": 441
+    },
+    {
+      "epoch": 1.3957894736842105,
+      "grad_norm": 0.7377170920372009,
+      "learning_rate": 6.443160905221188e-06,
+      "loss": 0.6848,
+      "step": 442
+    },
+    {
+      "epoch": 1.3989473684210525,
+      "grad_norm": 0.6914153099060059,
+      "learning_rate": 6.425519934216204e-06,
+      "loss": 0.7005,
+      "step": 443
+    },
+    {
+      "epoch": 1.4021052631578947,
+      "grad_norm": 0.6577550172805786,
+      "learning_rate": 6.407859626861734e-06,
+      "loss": 0.6872,
+      "step": 444
+    },
+    {
+      "epoch": 1.4052631578947368,
+      "grad_norm": 0.7401857972145081,
+      "learning_rate": 6.390180222709583e-06,
+      "loss": 0.6642,
+      "step": 445
+    },
+    {
+      "epoch": 1.408421052631579,
+      "grad_norm": 0.653255820274353,
+      "learning_rate": 6.372481961570597e-06,
+      "loss": 0.6641,
+      "step": 446
+    },
+    {
+      "epoch": 1.411578947368421,
+      "grad_norm": 0.5870925784111023,
+      "learning_rate": 6.3547650835114014e-06,
+      "loss": 0.6783,
+      "step": 447
+    },
+    {
+      "epoch": 1.4147368421052633,
+      "grad_norm": 0.7446110248565674,
+      "learning_rate": 6.337029828851151e-06,
+      "loss": 0.6965,
+      "step": 448
+    },
+    {
+      "epoch": 1.4178947368421053,
+      "grad_norm": 0.7110916972160339,
+      "learning_rate": 6.319276438158271e-06,
+      "loss": 0.6985,
+      "step": 449
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.6280571222305298,
+      "learning_rate": 6.301505152247185e-06,
+      "loss": 0.6855,
+      "step": 450
+    },
+    {
+      "epoch": 1.4242105263157896,
+      "grad_norm": 0.6359561681747437,
+      "learning_rate": 6.283716212175062e-06,
+      "loss": 0.6676,
+      "step": 451
+    },
+    {
+      "epoch": 1.4273684210526316,
+      "grad_norm": 0.670707106590271,
+      "learning_rate": 6.265909859238536e-06,
+      "loss": 0.6898,
+      "step": 452
+    },
+    {
+      "epoch": 1.4305263157894736,
+      "grad_norm": 0.6176190376281738,
+      "learning_rate": 6.248086334970435e-06,
+      "loss": 0.6481,
+      "step": 453
+    },
+    {
+      "epoch": 1.433684210526316,
+      "grad_norm": 0.6055241823196411,
+      "learning_rate": 6.230245881136509e-06,
+      "loss": 0.6382,
+      "step": 454
+    },
+    {
+      "epoch": 1.436842105263158,
+      "grad_norm": 0.6523144245147705,
+      "learning_rate": 6.2123887397321456e-06,
+      "loss": 0.7032,
+      "step": 455
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6355512142181396,
+      "learning_rate": 6.194515152979093e-06,
+      "loss": 0.6795,
+      "step": 456
+    },
+    {
+      "epoch": 1.4431578947368422,
+      "grad_norm": 0.5738744735717773,
+      "learning_rate": 6.176625363322164e-06,
+      "loss": 0.7088,
+      "step": 457
+    },
+    {
+      "epoch": 1.4463157894736842,
+      "grad_norm": 0.5846467018127441,
+      "learning_rate": 6.158719613425964e-06,
+      "loss": 0.7205,
+      "step": 458
+    },
+    {
+      "epoch": 1.4494736842105262,
+      "grad_norm": 0.6887890696525574,
+      "learning_rate": 6.140798146171581e-06,
+      "loss": 0.7071,
+      "step": 459
+    },
+    {
+      "epoch": 1.4526315789473685,
+      "grad_norm": 0.6898921132087708,
+      "learning_rate": 6.122861204653304e-06,
+      "loss": 0.6667,
+      "step": 460
+    },
+    {
+      "epoch": 1.4557894736842105,
+      "grad_norm": 0.6479648351669312,
+      "learning_rate": 6.104909032175323e-06,
+      "loss": 0.6884,
+      "step": 461
+    },
+    {
+      "epoch": 1.4589473684210525,
+      "grad_norm": 0.7151727676391602,
+      "learning_rate": 6.086941872248424e-06,
+      "loss": 0.6796,
+      "step": 462
+    },
+    {
+      "epoch": 1.4621052631578948,
+      "grad_norm": 0.649958074092865,
+      "learning_rate": 6.068959968586689e-06,
+      "loss": 0.7084,
+      "step": 463
+    },
+    {
+      "epoch": 1.4652631578947368,
+      "grad_norm": 0.590797483921051,
+      "learning_rate": 6.050963565104191e-06,
+      "loss": 0.6626,
+      "step": 464
+    },
+    {
+      "epoch": 1.4684210526315788,
+      "grad_norm": 0.6719523072242737,
+      "learning_rate": 6.032952905911686e-06,
+      "loss": 0.6946,
+      "step": 465
+    },
+    {
+      "epoch": 1.471578947368421,
+      "grad_norm": 0.6793899536132812,
+      "learning_rate": 6.014928235313301e-06,
+      "loss": 0.6863,
+      "step": 466
+    },
+    {
+      "epoch": 1.4747368421052631,
+      "grad_norm": 0.6513259410858154,
+      "learning_rate": 5.996889797803214e-06,
+      "loss": 0.7068,
+      "step": 467
+    },
+    {
+      "epoch": 1.4778947368421052,
+      "grad_norm": 0.5904105305671692,
+      "learning_rate": 5.978837838062348e-06,
+      "loss": 0.6878,
+      "step": 468
+    },
+    {
+      "epoch": 1.4810526315789474,
+      "grad_norm": 0.6193195581436157,
+      "learning_rate": 5.9607726009550494e-06,
+      "loss": 0.6824,
+      "step": 469
+    },
+    {
+      "epoch": 1.4842105263157894,
+      "grad_norm": 0.6615768074989319,
+      "learning_rate": 5.942694331525758e-06,
+      "loss": 0.6957,
+      "step": 470
+    },
+    {
+      "epoch": 1.4873684210526315,
+      "grad_norm": 0.5977884531021118,
+      "learning_rate": 5.924603274995693e-06,
+      "loss": 0.6732,
+      "step": 471
+    },
+    {
+      "epoch": 1.4905263157894737,
+      "grad_norm": 0.6094414591789246,
+      "learning_rate": 5.906499676759524e-06,
+      "loss": 0.6896,
+      "step": 472
+    },
+    {
+      "epoch": 1.4936842105263157,
+      "grad_norm": 0.5905715227127075,
+      "learning_rate": 5.88838378238204e-06,
+      "loss": 0.6964,
+      "step": 473
+    },
+    {
+      "epoch": 1.496842105263158,
+      "grad_norm": 0.6121339797973633,
+      "learning_rate": 5.8702558375948206e-06,
+      "loss": 0.6704,
+      "step": 474
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5825003385543823,
+      "learning_rate": 5.852116088292901e-06,
+      "loss": 0.6957,
+      "step": 475
+    },
+    {
+      "epoch": 1.5031578947368422,
+      "grad_norm": 0.6180065870285034,
+      "learning_rate": 5.8339647805314404e-06,
+      "loss": 0.7134,
+      "step": 476
+    },
+    {
+      "epoch": 1.506315789473684,
+      "grad_norm": 0.6621575355529785,
+      "learning_rate": 5.815802160522379e-06,
+      "loss": 0.6579,
+      "step": 477
+    },
+    {
+      "epoch": 1.5094736842105263,
+      "grad_norm": 0.6555248498916626,
+      "learning_rate": 5.797628474631102e-06,
+      "loss": 0.6424,
+      "step": 478
+    },
+    {
+      "epoch": 1.5126315789473685,
+      "grad_norm": 0.6568667888641357,
+      "learning_rate": 5.7794439693730975e-06,
+      "loss": 0.6965,
+      "step": 479
+    },
+    {
+      "epoch": 1.5157894736842106,
+      "grad_norm": 0.6549687385559082,
+      "learning_rate": 5.761248891410613e-06,
+      "loss": 0.7017,
+      "step": 480
+    },
+    {
+      "epoch": 1.5189473684210526,
+      "grad_norm": 0.6165953874588013,
+      "learning_rate": 5.743043487549306e-06,
+      "loss": 0.719,
+      "step": 481
+    },
+    {
+      "epoch": 1.5221052631578948,
+      "grad_norm": 0.6918132901191711,
+      "learning_rate": 5.7248280047348995e-06,
+      "loss": 0.6964,
+      "step": 482
+    },
+    {
+      "epoch": 1.5252631578947369,
+      "grad_norm": 0.6982928514480591,
+      "learning_rate": 5.706602690049832e-06,
+      "loss": 0.6933,
+      "step": 483
+    },
+    {
+      "epoch": 1.528421052631579,
+      "grad_norm": 0.6259163618087769,
+      "learning_rate": 5.688367790709909e-06,
+      "loss": 0.7202,
+      "step": 484
+    },
+    {
+      "epoch": 1.5315789473684212,
+      "grad_norm": 0.7041128873825073,
+      "learning_rate": 5.6701235540609405e-06,
+      "loss": 0.7115,
+      "step": 485
+    },
+    {
+      "epoch": 1.5347368421052632,
+      "grad_norm": 0.6209255456924438,
+      "learning_rate": 5.651870227575391e-06,
+      "loss": 0.6945,
+      "step": 486
+    },
+    {
+      "epoch": 1.5378947368421052,
+      "grad_norm": 0.6310893893241882,
+      "learning_rate": 5.633608058849033e-06,
+      "loss": 0.6985,
+      "step": 487
+    },
+    {
+      "epoch": 1.5410526315789475,
+      "grad_norm": 0.6265409588813782,
+      "learning_rate": 5.61533729559757e-06,
+      "loss": 0.7472,
+      "step": 488
+    },
+    {
+      "epoch": 1.5442105263157895,
+      "grad_norm": 0.621372640132904,
+      "learning_rate": 5.5970581856532864e-06,
+      "loss": 0.6869,
+      "step": 489
+    },
+    {
+      "epoch": 1.5473684210526315,
+      "grad_norm": 0.6584547758102417,
+      "learning_rate": 5.578770976961685e-06,
+      "loss": 0.6797,
+      "step": 490
+    },
+    {
+      "epoch": 1.5505263157894738,
+      "grad_norm": 0.6084906458854675,
+      "learning_rate": 5.560475917578129e-06,
+      "loss": 0.6897,
+      "step": 491
+    },
+    {
+      "epoch": 1.5536842105263158,
+      "grad_norm": 0.6554166674613953,
+      "learning_rate": 5.542173255664463e-06,
+      "loss": 0.6844,
+      "step": 492
+    },
+    {
+      "epoch": 1.5568421052631578,
+      "grad_norm": 0.5814923048019409,
+      "learning_rate": 5.523863239485661e-06,
+      "loss": 0.7098,
+      "step": 493
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6252988576889038,
+      "learning_rate": 5.505546117406449e-06,
+      "loss": 0.7227,
+      "step": 494
+    },
+    {
+      "epoch": 1.563157894736842,
+      "grad_norm": 0.6618260145187378,
+      "learning_rate": 5.487222137887949e-06,
+      "loss": 0.6874,
+      "step": 495
+    },
+    {
+      "epoch": 1.566315789473684,
+      "grad_norm": 0.6781257390975952,
+      "learning_rate": 5.4688915494842886e-06,
+      "loss": 0.6582,
+      "step": 496
+    },
+    {
+      "epoch": 1.5694736842105264,
+      "grad_norm": 0.7620246410369873,
+      "learning_rate": 5.450554600839251e-06,
+      "loss": 0.6982,
+      "step": 497
+    },
+    {
+      "epoch": 1.5726315789473684,
+      "grad_norm": 0.646678626537323,
+      "learning_rate": 5.432211540682887e-06,
+      "loss": 0.6742,
+      "step": 498
+    },
+    {
+      "epoch": 1.5757894736842104,
+      "grad_norm": 0.7081868648529053,
+      "learning_rate": 5.413862617828147e-06,
+      "loss": 0.656,
+      "step": 499
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.6851574182510376,
+      "learning_rate": 5.395508081167506e-06,
+      "loss": 0.6885,
+      "step": 500
+    },
+    {
+      "epoch": 1.582105263157895,
+      "grad_norm": 0.679471492767334,
+      "learning_rate": 5.37714817966959e-06,
+      "loss": 0.6661,
+      "step": 501
+    },
+    {
+      "epoch": 1.5852631578947367,
+      "grad_norm": 0.6246220469474792,
+      "learning_rate": 5.35878316237579e-06,
+      "loss": 0.6632,
+      "step": 502
+    },
+    {
+      "epoch": 1.588421052631579,
+      "grad_norm": 0.6367597579956055,
+      "learning_rate": 5.340413278396896e-06,
+      "loss": 0.6997,
+      "step": 503
+    },
+    {
+      "epoch": 1.5915789473684212,
+      "grad_norm": 0.7005538940429688,
+      "learning_rate": 5.322038776909705e-06,
+      "loss": 0.7041,
+      "step": 504
+    },
+    {
+      "epoch": 1.594736842105263,
+      "grad_norm": 0.6432095766067505,
+      "learning_rate": 5.303659907153654e-06,
+      "loss": 0.6719,
+      "step": 505
+    },
+    {
+      "epoch": 1.5978947368421053,
+      "grad_norm": 0.5566903352737427,
+      "learning_rate": 5.285276918427432e-06,
+      "loss": 0.6733,
+      "step": 506
+    },
+    {
+      "epoch": 1.6010526315789475,
+      "grad_norm": 0.5807555317878723,
+      "learning_rate": 5.2668900600855955e-06,
+      "loss": 0.7036,
+      "step": 507
+    },
+    {
+      "epoch": 1.6042105263157893,
+      "grad_norm": 0.5966797471046448,
+      "learning_rate": 5.248499581535193e-06,
+      "loss": 0.682,
+      "step": 508
+    },
+    {
+      "epoch": 1.6073684210526316,
+      "grad_norm": 0.646507740020752,
+      "learning_rate": 5.2301057322323786e-06,
+      "loss": 0.7059,
+      "step": 509
+    },
+    {
+      "epoch": 1.6105263157894738,
+      "grad_norm": 0.6112558841705322,
+      "learning_rate": 5.211708761679031e-06,
+      "loss": 0.6732,
+      "step": 510
+    },
+    {
+      "epoch": 1.6136842105263158,
+      "grad_norm": 0.6396105885505676,
+      "learning_rate": 5.193308919419363e-06,
+      "loss": 0.7091,
+      "step": 511
+    },
+    {
+      "epoch": 1.6168421052631579,
+      "grad_norm": 0.678657054901123,
+      "learning_rate": 5.1749064550365414e-06,
+      "loss": 0.6975,
+      "step": 512
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6183331608772278,
+      "learning_rate": 5.156501618149301e-06,
+      "loss": 0.6832,
+      "step": 513
+    },
+    {
+      "epoch": 1.6231578947368421,
+      "grad_norm": 0.6091780066490173,
+      "learning_rate": 5.13809465840856e-06,
+      "loss": 0.6911,
+      "step": 514
+    },
+    {
+      "epoch": 1.6263157894736842,
+      "grad_norm": 0.6304209232330322,
+      "learning_rate": 5.11968582549403e-06,
+      "loss": 0.6461,
+      "step": 515
+    },
+    {
+      "epoch": 1.6294736842105264,
+      "grad_norm": 0.6831485629081726,
+      "learning_rate": 5.10127536911083e-06,
+      "loss": 0.6757,
+      "step": 516
+    },
+    {
+      "epoch": 1.6326315789473684,
+      "grad_norm": 0.6476657390594482,
+      "learning_rate": 5.082863538986103e-06,
+      "loss": 0.6935,
+      "step": 517
+    },
+    {
+      "epoch": 1.6357894736842105,
+      "grad_norm": 0.6213446259498596,
+      "learning_rate": 5.064450584865624e-06,
+      "loss": 0.668,
+      "step": 518
+    },
+    {
+      "epoch": 1.6389473684210527,
+      "grad_norm": 0.6051395535469055,
+      "learning_rate": 5.046036756510417e-06,
+      "loss": 0.6926,
+      "step": 519
+    },
+    {
+      "epoch": 1.6421052631578947,
+      "grad_norm": 0.6724497079849243,
+      "learning_rate": 5.027622303693363e-06,
+      "loss": 0.6912,
+      "step": 520
+    },
+    {
+      "epoch": 1.6452631578947368,
+      "grad_norm": 0.6436070799827576,
+      "learning_rate": 5.0092074761958085e-06,
+      "loss": 0.6791,
+      "step": 521
+    },
+    {
+      "epoch": 1.648421052631579,
+      "grad_norm": 0.602382242679596,
+      "learning_rate": 4.990792523804192e-06,
+      "loss": 0.7104,
+      "step": 522
+    },
+    {
+      "epoch": 1.651578947368421,
+      "grad_norm": 0.6710056662559509,
+      "learning_rate": 4.972377696306639e-06,
+      "loss": 0.6412,
+      "step": 523
+    },
+    {
+      "epoch": 1.654736842105263,
+      "grad_norm": 0.6339631676673889,
+      "learning_rate": 4.953963243489583e-06,
+      "loss": 0.696,
+      "step": 524
+    },
+    {
+      "epoch": 1.6578947368421053,
+      "grad_norm": 0.6232373118400574,
+      "learning_rate": 4.935549415134376e-06,
+      "loss": 0.6824,
+      "step": 525
+    },
+    {
+      "epoch": 1.6610526315789473,
+      "grad_norm": 0.5875207781791687,
+      "learning_rate": 4.9171364610139e-06,
+      "loss": 0.7114,
+      "step": 526
+    },
+    {
+      "epoch": 1.6642105263157894,
+      "grad_norm": 0.6042991876602173,
+      "learning_rate": 4.898724630889172e-06,
+      "loss": 0.6792,
+      "step": 527
+    },
+    {
+      "epoch": 1.6673684210526316,
+      "grad_norm": 0.6709006428718567,
+      "learning_rate": 4.880314174505972e-06,
+      "loss": 0.6597,
+      "step": 528
+    },
+    {
+      "epoch": 1.6705263157894736,
+      "grad_norm": 0.6271777153015137,
+      "learning_rate": 4.861905341591442e-06,
+      "loss": 0.6763,
+      "step": 529
+    },
+    {
+      "epoch": 1.6736842105263157,
+      "grad_norm": 0.6083865165710449,
+      "learning_rate": 4.843498381850701e-06,
+      "loss": 0.6882,
+      "step": 530
+    },
+    {
+      "epoch": 1.676842105263158,
+      "grad_norm": 0.544291079044342,
+      "learning_rate": 4.82509354496346e-06,
+      "loss": 0.705,
+      "step": 531
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.6035980582237244,
+      "learning_rate": 4.8066910805806384e-06,
+      "loss": 0.6771,
+      "step": 532
+    },
+    {
+      "epoch": 1.683157894736842,
+      "grad_norm": 0.6175979971885681,
+      "learning_rate": 4.78829123832097e-06,
+      "loss": 0.7164,
+      "step": 533
+    },
+    {
+      "epoch": 1.6863157894736842,
+      "grad_norm": 0.6440612077713013,
+      "learning_rate": 4.769894267767621e-06,
+      "loss": 0.6812,
+      "step": 534
+    },
+    {
+      "epoch": 1.6894736842105265,
+      "grad_norm": 0.6156215071678162,
+      "learning_rate": 4.751500418464809e-06,
+      "loss": 0.6916,
+      "step": 535
+    },
+    {
+      "epoch": 1.6926315789473683,
+      "grad_norm": 0.5670990943908691,
+      "learning_rate": 4.733109939914407e-06,
+      "loss": 0.6901,
+      "step": 536
+    },
+    {
+      "epoch": 1.6957894736842105,
+      "grad_norm": 0.6149737238883972,
+      "learning_rate": 4.714723081572571e-06,
+      "loss": 0.6412,
+      "step": 537
+    },
+    {
+      "epoch": 1.6989473684210528,
+      "grad_norm": 0.6131264567375183,
+      "learning_rate": 4.696340092846347e-06,
+      "loss": 0.6881,
+      "step": 538
+    },
+    {
+      "epoch": 1.7021052631578948,
+      "grad_norm": 0.5654208660125732,
+      "learning_rate": 4.677961223090297e-06,
+      "loss": 0.7101,
+      "step": 539
+    },
+    {
+      "epoch": 1.7052631578947368,
+      "grad_norm": 0.6652705073356628,
+      "learning_rate": 4.659586721603107e-06,
+      "loss": 0.7,
+      "step": 540
+    },
+    {
+      "epoch": 1.708421052631579,
+      "grad_norm": 0.6081032752990723,
+      "learning_rate": 4.641216837624211e-06,
+      "loss": 0.7095,
+      "step": 541
+    },
+    {
+      "epoch": 1.711578947368421,
+      "grad_norm": 0.5956637263298035,
+      "learning_rate": 4.622851820330412e-06,
+      "loss": 0.6858,
+      "step": 542
+    },
+    {
+      "epoch": 1.7147368421052631,
+      "grad_norm": 0.6079306602478027,
+      "learning_rate": 4.604491918832494e-06,
+      "loss": 0.6929,
+      "step": 543
+    },
+    {
+      "epoch": 1.7178947368421054,
+      "grad_norm": 0.5816709399223328,
+      "learning_rate": 4.586137382171856e-06,
+      "loss": 0.6681,
+      "step": 544
+    },
+    {
+      "epoch": 1.7210526315789474,
+      "grad_norm": 0.6434882879257202,
+      "learning_rate": 4.567788459317116e-06,
+      "loss": 0.6812,
+      "step": 545
+    },
+    {
+      "epoch": 1.7242105263157894,
+      "grad_norm": 0.6063170433044434,
+      "learning_rate": 4.54944539916075e-06,
+      "loss": 0.6832,
+      "step": 546
+    },
+    {
+      "epoch": 1.7273684210526317,
+      "grad_norm": 0.6341161727905273,
+      "learning_rate": 4.531108450515712e-06,
+      "loss": 0.692,
+      "step": 547
+    },
+    {
+      "epoch": 1.7305263157894737,
+      "grad_norm": 0.6072157621383667,
+      "learning_rate": 4.512777862112053e-06,
+      "loss": 0.6582,
+      "step": 548
+    },
+    {
+      "epoch": 1.7336842105263157,
+      "grad_norm": 0.5953732132911682,
+      "learning_rate": 4.494453882593552e-06,
+      "loss": 0.663,
+      "step": 549
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.6301018595695496,
+      "learning_rate": 4.476136760514341e-06,
+      "loss": 0.6686,
+      "step": 550
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.628161609172821,
+      "learning_rate": 4.457826744335538e-06,
+      "loss": 0.6909,
+      "step": 551
+    },
+    {
+      "epoch": 1.743157894736842,
+      "grad_norm": 0.5877235531806946,
+      "learning_rate": 4.439524082421872e-06,
+      "loss": 0.6515,
+      "step": 552
+    },
+    {
+      "epoch": 1.7463157894736843,
+      "grad_norm": 0.6606840491294861,
+      "learning_rate": 4.421229023038316e-06,
+      "loss": 0.6738,
+      "step": 553
+    },
+    {
+      "epoch": 1.7494736842105263,
+      "grad_norm": 0.5608782768249512,
+      "learning_rate": 4.402941814346716e-06,
+      "loss": 0.6991,
+      "step": 554
+    },
+    {
+      "epoch": 1.7526315789473683,
+      "grad_norm": 0.5692189335823059,
+      "learning_rate": 4.384662704402433e-06,
+      "loss": 0.6547,
+      "step": 555
+    },
+    {
+      "epoch": 1.7557894736842106,
+      "grad_norm": 0.589918315410614,
+      "learning_rate": 4.366391941150969e-06,
+      "loss": 0.691,
+      "step": 556
+    },
+    {
+      "epoch": 1.7589473684210526,
+      "grad_norm": 0.5952433943748474,
+      "learning_rate": 4.34812977242461e-06,
+      "loss": 0.6866,
+      "step": 557
+    },
+    {
+      "epoch": 1.7621052631578946,
+      "grad_norm": 0.6127429008483887,
+      "learning_rate": 4.329876445939062e-06,
+      "loss": 0.7041,
+      "step": 558
+    },
+    {
+      "epoch": 1.7652631578947369,
+      "grad_norm": 0.6347705125808716,
+      "learning_rate": 4.3116322092900925e-06,
+      "loss": 0.6565,
+      "step": 559
+    },
+    {
+      "epoch": 1.768421052631579,
+      "grad_norm": 0.5670740604400635,
+      "learning_rate": 4.293397309950168e-06,
+      "loss": 0.7044,
+      "step": 560
+    },
+    {
+      "epoch": 1.771578947368421,
+      "grad_norm": 0.6550533175468445,
+      "learning_rate": 4.275171995265101e-06,
+      "loss": 0.6867,
+      "step": 561
+    },
+    {
+      "epoch": 1.7747368421052632,
+      "grad_norm": 0.5752691030502319,
+      "learning_rate": 4.256956512450697e-06,
+      "loss": 0.7015,
+      "step": 562
+    },
+    {
+      "epoch": 1.7778947368421054,
+      "grad_norm": 0.6235054731369019,
+      "learning_rate": 4.238751108589389e-06,
+      "loss": 0.6954,
+      "step": 563
+    },
+    {
+      "epoch": 1.7810526315789472,
+      "grad_norm": 0.6959113478660583,
+      "learning_rate": 4.220556030626904e-06,
+      "loss": 0.6945,
+      "step": 564
+    },
+    {
+      "epoch": 1.7842105263157895,
+      "grad_norm": 0.5536004304885864,
+      "learning_rate": 4.202371525368899e-06,
+      "loss": 0.6849,
+      "step": 565
+    },
+    {
+      "epoch": 1.7873684210526317,
+      "grad_norm": 0.5802682042121887,
+      "learning_rate": 4.184197839477622e-06,
+      "loss": 0.6898,
+      "step": 566
+    },
+    {
+      "epoch": 1.7905263157894735,
+      "grad_norm": 0.6382439136505127,
+      "learning_rate": 4.166035219468561e-06,
+      "loss": 0.6784,
+      "step": 567
+    },
+    {
+      "epoch": 1.7936842105263158,
+      "grad_norm": 0.5798212289810181,
+      "learning_rate": 4.1478839117071e-06,
+      "loss": 0.7353,
+      "step": 568
+    },
+    {
+      "epoch": 1.796842105263158,
+      "grad_norm": 0.6376686096191406,
+      "learning_rate": 4.12974416240518e-06,
+      "loss": 0.665,
+      "step": 569
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.566200852394104,
+      "learning_rate": 4.11161621761796e-06,
+      "loss": 0.6689,
+      "step": 570
+    },
+    {
+      "epoch": 1.803157894736842,
+      "grad_norm": 0.6278012990951538,
+      "learning_rate": 4.093500323240479e-06,
+      "loss": 0.6716,
+      "step": 571
+    },
+    {
+      "epoch": 1.8063157894736843,
+      "grad_norm": 0.5599969625473022,
+      "learning_rate": 4.075396725004308e-06,
+      "loss": 0.6789,
+      "step": 572
+    },
+    {
+      "epoch": 1.8094736842105263,
+      "grad_norm": 0.6141947507858276,
+      "learning_rate": 4.057305668474244e-06,
+      "loss": 0.6696,
+      "step": 573
+    },
+    {
+      "epoch": 1.8126315789473684,
+      "grad_norm": 0.6866646409034729,
+      "learning_rate": 4.039227399044952e-06,
+      "loss": 0.7151,
+      "step": 574
+    },
+    {
+      "epoch": 1.8157894736842106,
+      "grad_norm": 0.5738588571548462,
+      "learning_rate": 4.021162161937653e-06,
+      "loss": 0.6942,
+      "step": 575
+    },
+    {
+      "epoch": 1.8189473684210526,
+      "grad_norm": 0.6379348039627075,
+      "learning_rate": 4.003110202196787e-06,
+      "loss": 0.6812,
+      "step": 576
+    },
+    {
+      "epoch": 1.8221052631578947,
+      "grad_norm": 0.6298931837081909,
+      "learning_rate": 3.985071764686701e-06,
+      "loss": 0.6711,
+      "step": 577
+    },
+    {
+      "epoch": 1.825263157894737,
+      "grad_norm": 0.6079388856887817,
+      "learning_rate": 3.9670470940883144e-06,
+      "loss": 0.6989,
+      "step": 578
+    },
+    {
+      "epoch": 1.828421052631579,
+      "grad_norm": 0.5470216870307922,
+      "learning_rate": 3.94903643489581e-06,
+      "loss": 0.7104,
+      "step": 579
+    },
+    {
+      "epoch": 1.831578947368421,
+      "grad_norm": 0.5332590341567993,
+      "learning_rate": 3.931040031413313e-06,
+      "loss": 0.6709,
+      "step": 580
+    },
+    {
+      "epoch": 1.8347368421052632,
+      "grad_norm": 0.5865632891654968,
+      "learning_rate": 3.913058127751578e-06,
+      "loss": 0.6886,
+      "step": 581
+    },
+    {
+      "epoch": 1.8378947368421052,
+      "grad_norm": 0.55133056640625,
+      "learning_rate": 3.895090967824678e-06,
+      "loss": 0.6543,
+      "step": 582
+    },
+    {
+      "epoch": 1.8410526315789473,
+      "grad_norm": 0.5989318490028381,
+      "learning_rate": 3.877138795346697e-06,
+      "loss": 0.64,
+      "step": 583
+    },
+    {
+      "epoch": 1.8442105263157895,
+      "grad_norm": 0.7691587805747986,
+      "learning_rate": 3.85920185382842e-06,
+      "loss": 0.6826,
+      "step": 584
+    },
+    {
+      "epoch": 1.8473684210526315,
+      "grad_norm": 0.6429767608642578,
+      "learning_rate": 3.841280386574037e-06,
+      "loss": 0.7013,
+      "step": 585
+    },
+    {
+      "epoch": 1.8505263157894736,
+      "grad_norm": 0.6208791732788086,
+      "learning_rate": 3.823374636677837e-06,
+      "loss": 0.6459,
+      "step": 586
+    },
+    {
+      "epoch": 1.8536842105263158,
+      "grad_norm": 0.5830709934234619,
+      "learning_rate": 3.8054848470209094e-06,
+      "loss": 0.678,
+      "step": 587
+    },
+    {
+      "epoch": 1.8568421052631578,
+      "grad_norm": 0.7179412245750427,
+      "learning_rate": 3.7876112602678544e-06,
+      "loss": 0.7003,
+      "step": 588
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 0.72615647315979,
+      "learning_rate": 3.7697541188634934e-06,
+      "loss": 0.6505,
+      "step": 589
+    },
+    {
+      "epoch": 1.8631578947368421,
+      "grad_norm": 0.6515791416168213,
+      "learning_rate": 3.7519136650295673e-06,
+      "loss": 0.7111,
+      "step": 590
+    },
+    {
+      "epoch": 1.8663157894736844,
+      "grad_norm": 0.6129830479621887,
+      "learning_rate": 3.734090140761466e-06,
+      "loss": 0.6769,
+      "step": 591
+    },
+    {
+      "epoch": 1.8694736842105262,
+      "grad_norm": 0.6099033951759338,
+      "learning_rate": 3.716283787824939e-06,
+      "loss": 0.6742,
+      "step": 592
+    },
+    {
+      "epoch": 1.8726315789473684,
+      "grad_norm": 0.6122466325759888,
+      "learning_rate": 3.698494847752816e-06,
+      "loss": 0.6786,
+      "step": 593
+    },
+    {
+      "epoch": 1.8757894736842107,
+      "grad_norm": 0.6633381843566895,
+      "learning_rate": 3.6807235618417314e-06,
+      "loss": 0.6878,
+      "step": 594
+    },
+    {
+      "epoch": 1.8789473684210525,
+      "grad_norm": 0.669661819934845,
+      "learning_rate": 3.6629701711488485e-06,
+      "loss": 0.6794,
+      "step": 595
+    },
+    {
+      "epoch": 1.8821052631578947,
+      "grad_norm": 0.5703924894332886,
+      "learning_rate": 3.645234916488599e-06,
+      "loss": 0.6814,
+      "step": 596
+    },
+    {
+      "epoch": 1.885263157894737,
+      "grad_norm": 0.6017104387283325,
+      "learning_rate": 3.6275180384294033e-06,
+      "loss": 0.6442,
+      "step": 597
+    },
+    {
+      "epoch": 1.888421052631579,
+      "grad_norm": 0.5789917707443237,
+      "learning_rate": 3.609819777290418e-06,
+      "loss": 0.6669,
+      "step": 598
+    },
+    {
+      "epoch": 1.891578947368421,
+      "grad_norm": 0.6235988140106201,
+      "learning_rate": 3.5921403731382685e-06,
+      "loss": 0.636,
+      "step": 599
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.6223317384719849,
+      "learning_rate": 3.5744800657837984e-06,
+      "loss": 0.6911,
+      "step": 600
+    },
+    {
+      "epoch": 1.8978947368421053,
+      "grad_norm": 0.6065788269042969,
+      "learning_rate": 3.556839094778814e-06,
+      "loss": 0.6495,
+      "step": 601
+    },
+    {
+      "epoch": 1.9010526315789473,
+      "grad_norm": 0.6049926280975342,
+      "learning_rate": 3.5392176994128357e-06,
+      "loss": 0.7152,
+      "step": 602
+    },
+    {
+      "epoch": 1.9042105263157896,
+      "grad_norm": 0.5944041609764099,
+      "learning_rate": 3.5216161187098497e-06,
+      "loss": 0.6835,
+      "step": 603
+    },
+    {
+      "epoch": 1.9073684210526316,
+      "grad_norm": 0.6229305863380432,
+      "learning_rate": 3.504034591425071e-06,
+      "loss": 0.674,
+      "step": 604
+    },
+    {
+      "epoch": 1.9105263157894736,
+      "grad_norm": 0.6037994623184204,
+      "learning_rate": 3.4864733560416998e-06,
+      "loss": 0.6526,
+      "step": 605
+    },
+    {
+      "epoch": 1.9136842105263159,
+      "grad_norm": 0.6101519465446472,
+      "learning_rate": 3.468932650767689e-06,
+      "loss": 0.6693,
+      "step": 606
+    },
+    {
+      "epoch": 1.916842105263158,
+      "grad_norm": 0.5631049871444702,
+      "learning_rate": 3.4514127135325105e-06,
+      "loss": 0.6967,
+      "step": 607
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5985934734344482,
+      "learning_rate": 3.433913781983932e-06,
+      "loss": 0.6955,
+      "step": 608
+    },
+    {
+      "epoch": 1.9231578947368422,
+      "grad_norm": 0.5818445086479187,
+      "learning_rate": 3.4164360934847912e-06,
+      "loss": 0.6642,
+      "step": 609
+    },
+    {
+      "epoch": 1.9263157894736842,
+      "grad_norm": 0.6571829319000244,
+      "learning_rate": 3.3989798851097744e-06,
+      "loss": 0.6848,
+      "step": 610
+    },
+    {
+      "epoch": 1.9294736842105262,
+      "grad_norm": 0.5725364685058594,
+      "learning_rate": 3.381545393642205e-06,
+      "loss": 0.6863,
+      "step": 611
+    },
+    {
+      "epoch": 1.9326315789473685,
+      "grad_norm": 0.6822765469551086,
+      "learning_rate": 3.3641328555708286e-06,
+      "loss": 0.7075,
+      "step": 612
+    },
+    {
+      "epoch": 1.9357894736842105,
+      "grad_norm": 0.6487634778022766,
+      "learning_rate": 3.3467425070866034e-06,
+      "loss": 0.675,
+      "step": 613
+    },
+    {
+      "epoch": 1.9389473684210525,
+      "grad_norm": 0.6720603108406067,
+      "learning_rate": 3.3293745840795004e-06,
+      "loss": 0.6709,
+      "step": 614
+    },
+    {
+      "epoch": 1.9421052631578948,
+      "grad_norm": 0.5721080303192139,
+      "learning_rate": 3.312029322135306e-06,
+      "loss": 0.6994,
+      "step": 615
+    },
+    {
+      "epoch": 1.9452631578947368,
+      "grad_norm": 0.6841775178909302,
+      "learning_rate": 3.2947069565324134e-06,
+      "loss": 0.6535,
+      "step": 616
+    },
+    {
+      "epoch": 1.9484210526315788,
+      "grad_norm": 0.6342321038246155,
+      "learning_rate": 3.2774077222386465e-06,
+      "loss": 0.6453,
+      "step": 617
+    },
+    {
+      "epoch": 1.951578947368421,
+      "grad_norm": 0.667797863483429,
+      "learning_rate": 3.260131853908066e-06,
+      "loss": 0.7042,
+      "step": 618
+    },
+    {
+      "epoch": 1.954736842105263,
+      "grad_norm": 0.6030858755111694,
+      "learning_rate": 3.2428795858777873e-06,
+      "loss": 0.6576,
+      "step": 619
+    },
+    {
+      "epoch": 1.9578947368421051,
+      "grad_norm": 0.6035648584365845,
+      "learning_rate": 3.225651152164799e-06,
+      "loss": 0.7307,
+      "step": 620
+    },
+    {
+      "epoch": 1.9610526315789474,
+      "grad_norm": 0.6750203371047974,
+      "learning_rate": 3.208446786462791e-06,
+      "loss": 0.7058,
+      "step": 621
+    },
+    {
+      "epoch": 1.9642105263157896,
+      "grad_norm": 0.6008321046829224,
+      "learning_rate": 3.1912667221389892e-06,
+      "loss": 0.6851,
+      "step": 622
+    },
+    {
+      "epoch": 1.9673684210526314,
+      "grad_norm": 0.6270667910575867,
+      "learning_rate": 3.1741111922309797e-06,
+      "loss": 0.6851,
+      "step": 623
+    },
+    {
+      "epoch": 1.9705263157894737,
+      "grad_norm": 0.5572888255119324,
+      "learning_rate": 3.156980429443559e-06,
+      "loss": 0.6896,
+      "step": 624
+    },
+    {
+      "epoch": 1.973684210526316,
+      "grad_norm": 0.6308818459510803,
+      "learning_rate": 3.1398746661455647e-06,
+      "loss": 0.7103,
+      "step": 625
+    },
+    {
+      "epoch": 1.9768421052631577,
+      "grad_norm": 0.6266995668411255,
+      "learning_rate": 3.122794134366738e-06,
+      "loss": 0.6679,
+      "step": 626
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.5579388737678528,
+      "learning_rate": 3.105739065794565e-06,
+      "loss": 0.677,
+      "step": 627
+    },
+    {
+      "epoch": 1.9831578947368422,
+      "grad_norm": 0.6134604215621948,
+      "learning_rate": 3.0887096917711408e-06,
+      "loss": 0.6466,
+      "step": 628
+    },
+    {
+      "epoch": 1.9863157894736843,
+      "grad_norm": 0.6037647128105164,
+      "learning_rate": 3.071706243290026e-06,
+      "loss": 0.6906,
+      "step": 629
+    },
+    {
+      "epoch": 1.9894736842105263,
+      "grad_norm": 0.636805534362793,
+      "learning_rate": 3.0547289509931194e-06,
+      "loss": 0.6838,
+      "step": 630
+    },
+    {
+      "epoch": 1.9926315789473685,
+      "grad_norm": 0.6072667837142944,
+      "learning_rate": 3.0377780451675243e-06,
+      "loss": 0.6751,
+      "step": 631
+    },
+    {
+      "epoch": 1.9957894736842106,
+      "grad_norm": 0.5715926885604858,
+      "learning_rate": 3.020853755742428e-06,
+      "loss": 0.6666,
+      "step": 632
+    },
+    {
+      "epoch": 1.9989473684210526,
+      "grad_norm": 0.634506106376648,
+      "learning_rate": 3.0039563122859815e-06,
+      "loss": 0.706,
+      "step": 633
+    },
+    {
+      "epoch": 2.002105263157895,
+      "grad_norm": 0.6592942476272583,
+      "learning_rate": 2.9870859440021845e-06,
+      "loss": 0.6761,
+      "step": 634
+    },
+    {
+      "epoch": 2.0052631578947366,
+      "grad_norm": 0.7126978635787964,
+      "learning_rate": 2.970242879727778e-06,
+      "loss": 0.6233,
+      "step": 635
+    },
+    {
+      "epoch": 2.008421052631579,
+      "grad_norm": 0.5990601778030396,
+      "learning_rate": 2.953427347929142e-06,
+      "loss": 0.6303,
+      "step": 636
+    },
+    {
+      "epoch": 2.011578947368421,
+      "grad_norm": 0.5881582498550415,
+      "learning_rate": 2.936639576699194e-06,
+      "loss": 0.5866,
+      "step": 637
+    },
+    {
+      "epoch": 2.014736842105263,
+      "grad_norm": 0.6993144750595093,
+      "learning_rate": 2.9198797937542935e-06,
+      "loss": 0.6256,
+      "step": 638
+    },
+    {
+      "epoch": 2.017894736842105,
+      "grad_norm": 0.6200084686279297,
+      "learning_rate": 2.903148226431155e-06,
+      "loss": 0.6453,
+      "step": 639
+    },
+    {
+      "epoch": 2.0210526315789474,
+      "grad_norm": 0.6157467365264893,
+      "learning_rate": 2.8864451016837703e-06,
+      "loss": 0.606,
+      "step": 640
+    },
+    {
+      "epoch": 2.0242105263157897,
+      "grad_norm": 0.6459230780601501,
+      "learning_rate": 2.869770646080316e-06,
+      "loss": 0.6267,
+      "step": 641
+    },
+    {
+      "epoch": 2.0273684210526315,
+      "grad_norm": 0.7284769415855408,
+      "learning_rate": 2.853125085800096e-06,
+      "loss": 0.6036,
+      "step": 642
+    },
+    {
+      "epoch": 2.0305263157894737,
+      "grad_norm": 0.65320885181427,
+      "learning_rate": 2.836508646630457e-06,
+      "loss": 0.6798,
+      "step": 643
+    },
+    {
+      "epoch": 2.033684210526316,
+      "grad_norm": 0.6911686658859253,
+      "learning_rate": 2.8199215539637427e-06,
+      "loss": 0.6431,
+      "step": 644
+    },
+    {
+      "epoch": 2.036842105263158,
+      "grad_norm": 0.6208088994026184,
+      "learning_rate": 2.8033640327942235e-06,
+      "loss": 0.673,
+      "step": 645
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.6126965284347534,
+      "learning_rate": 2.786836307715056e-06,
+      "loss": 0.6152,
+      "step": 646
+    },
+    {
+      "epoch": 2.0431578947368423,
+      "grad_norm": 0.7363206148147583,
+      "learning_rate": 2.7703386029152246e-06,
+      "loss": 0.6492,
+      "step": 647
+    },
+    {
+      "epoch": 2.046315789473684,
+      "grad_norm": 0.6911171078681946,
+      "learning_rate": 2.753871142176506e-06,
+      "loss": 0.6315,
+      "step": 648
+    },
+    {
+      "epoch": 2.0494736842105263,
+      "grad_norm": 0.6493732929229736,
+      "learning_rate": 2.737434148870437e-06,
+      "loss": 0.6249,
+      "step": 649
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.6811050772666931,
+      "learning_rate": 2.7210278459552786e-06,
+      "loss": 0.6038,
+      "step": 650
+    },
+    {
+      "epoch": 2.0557894736842104,
+      "grad_norm": 0.6253058910369873,
+      "learning_rate": 2.704652455972997e-06,
+      "loss": 0.6369,
+      "step": 651
+    },
+    {
+      "epoch": 2.0589473684210526,
+      "grad_norm": 0.5684403777122498,
+      "learning_rate": 2.688308201046236e-06,
+      "loss": 0.6551,
+      "step": 652
+    },
+    {
+      "epoch": 2.062105263157895,
+      "grad_norm": 0.6057170629501343,
+      "learning_rate": 2.6719953028753214e-06,
+      "loss": 0.5905,
+      "step": 653
+    },
+    {
+      "epoch": 2.0652631578947367,
+      "grad_norm": 0.603461742401123,
+      "learning_rate": 2.655713982735234e-06,
+      "loss": 0.6049,
+      "step": 654
+    },
+    {
+      "epoch": 2.068421052631579,
+      "grad_norm": 0.5477900505065918,
+      "learning_rate": 2.6394644614726215e-06,
+      "loss": 0.6091,
+      "step": 655
+    },
+    {
+      "epoch": 2.071578947368421,
+      "grad_norm": 0.6187466382980347,
+      "learning_rate": 2.623246959502795e-06,
+      "loss": 0.6247,
+      "step": 656
+    },
+    {
+      "epoch": 2.074736842105263,
+      "grad_norm": 0.5489943623542786,
+      "learning_rate": 2.6070616968067446e-06,
+      "loss": 0.6522,
+      "step": 657
+    },
+    {
+      "epoch": 2.0778947368421052,
+      "grad_norm": 0.5987683534622192,
+      "learning_rate": 2.5909088929281534e-06,
+      "loss": 0.613,
+      "step": 658
+    },
+    {
+      "epoch": 2.0810526315789475,
+      "grad_norm": 0.5765239000320435,
+      "learning_rate": 2.574788766970418e-06,
+      "loss": 0.6277,
+      "step": 659
+    },
+    {
+      "epoch": 2.0842105263157893,
+      "grad_norm": 0.566733717918396,
+      "learning_rate": 2.55870153759368e-06,
+      "loss": 0.5948,
+      "step": 660
+    },
+    {
+      "epoch": 2.0873684210526315,
+      "grad_norm": 0.5783083438873291,
+      "learning_rate": 2.542647423011857e-06,
+      "loss": 0.6302,
+      "step": 661
+    },
+    {
+      "epoch": 2.090526315789474,
+      "grad_norm": 0.6213259696960449,
+      "learning_rate": 2.526626640989683e-06,
+      "loss": 0.607,
+      "step": 662
+    },
+    {
+      "epoch": 2.0936842105263156,
+      "grad_norm": 0.5684741735458374,
+      "learning_rate": 2.510639408839757e-06,
+      "loss": 0.6122,
+      "step": 663
+    },
+    {
+      "epoch": 2.096842105263158,
+      "grad_norm": 0.5746645927429199,
+      "learning_rate": 2.4946859434195904e-06,
+      "loss": 0.6336,
+      "step": 664
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.554362416267395,
+      "learning_rate": 2.478766461128672e-06,
+      "loss": 0.6155,
+      "step": 665
+    },
+    {
+      "epoch": 2.103157894736842,
+      "grad_norm": 0.5717495679855347,
+      "learning_rate": 2.4628811779055277e-06,
+      "loss": 0.6356,
+      "step": 666
+    },
+    {
+      "epoch": 2.106315789473684,
+      "grad_norm": 0.5523346662521362,
+      "learning_rate": 2.4470303092247926e-06,
+      "loss": 0.5855,
+      "step": 667
+    },
+    {
+      "epoch": 2.1094736842105264,
+      "grad_norm": 0.5319191813468933,
+      "learning_rate": 2.431214070094289e-06,
+      "loss": 0.5865,
+      "step": 668
+    },
+    {
+      "epoch": 2.1126315789473686,
+      "grad_norm": 0.52125483751297,
+      "learning_rate": 2.4154326750521084e-06,
+      "loss": 0.6155,
+      "step": 669
+    },
+    {
+      "epoch": 2.1157894736842104,
+      "grad_norm": 0.5644091963768005,
+      "learning_rate": 2.3996863381637046e-06,
+      "loss": 0.5797,
+      "step": 670
+    },
+    {
+      "epoch": 2.1189473684210527,
+      "grad_norm": 0.5224770307540894,
+      "learning_rate": 2.383975273018986e-06,
+      "loss": 0.6275,
+      "step": 671
+    },
+    {
+      "epoch": 2.122105263157895,
+      "grad_norm": 0.5387361645698547,
+      "learning_rate": 2.3682996927294216e-06,
+      "loss": 0.5888,
+      "step": 672
+    },
+    {
+      "epoch": 2.1252631578947367,
+      "grad_norm": 0.6090015172958374,
+      "learning_rate": 2.3526598099251473e-06,
+      "loss": 0.5825,
+      "step": 673
+    },
+    {
+      "epoch": 2.128421052631579,
+      "grad_norm": 0.5374984741210938,
+      "learning_rate": 2.3370558367520856e-06,
+      "loss": 0.6262,
+      "step": 674
+    },
+    {
+      "epoch": 2.1315789473684212,
+      "grad_norm": 0.581964373588562,
+      "learning_rate": 2.321487984869064e-06,
+      "loss": 0.5949,
+      "step": 675
+    },
+    {
+      "epoch": 2.134736842105263,
+      "grad_norm": 0.5389418601989746,
+      "learning_rate": 2.305956465444945e-06,
+      "loss": 0.6104,
+      "step": 676
+    },
+    {
+      "epoch": 2.1378947368421053,
+      "grad_norm": 0.5349326729774475,
+      "learning_rate": 2.290461489155768e-06,
+      "loss": 0.6828,
+      "step": 677
+    },
+    {
+      "epoch": 2.1410526315789475,
+      "grad_norm": 0.5772843360900879,
+      "learning_rate": 2.275003266181877e-06,
+      "loss": 0.5864,
+      "step": 678
+    },
+    {
+      "epoch": 2.1442105263157893,
+      "grad_norm": 0.6210830211639404,
+      "learning_rate": 2.2595820062050854e-06,
+      "loss": 0.6047,
+      "step": 679
+    },
+    {
+      "epoch": 2.1473684210526316,
+      "grad_norm": 0.5432112812995911,
+      "learning_rate": 2.2441979184058223e-06,
+      "loss": 0.6402,
+      "step": 680
+    },
+    {
+      "epoch": 2.150526315789474,
+      "grad_norm": 0.5438613891601562,
+      "learning_rate": 2.2288512114602986e-06,
+      "loss": 0.5905,
+      "step": 681
+    },
+    {
+      "epoch": 2.1536842105263156,
+      "grad_norm": 0.5581168532371521,
+      "learning_rate": 2.213542093537675e-06,
+      "loss": 0.5887,
+      "step": 682
+    },
+    {
+      "epoch": 2.156842105263158,
+      "grad_norm": 0.5446808338165283,
+      "learning_rate": 2.1982707722972383e-06,
+      "loss": 0.6573,
+      "step": 683
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6228784322738647,
+      "learning_rate": 2.1830374548855905e-06,
+      "loss": 0.6083,
+      "step": 684
+    },
+    {
+      "epoch": 2.163157894736842,
+      "grad_norm": 0.577155351638794,
+      "learning_rate": 2.167842347933826e-06,
+      "loss": 0.614,
+      "step": 685
+    },
+    {
+      "epoch": 2.166315789473684,
+      "grad_norm": 0.560425877571106,
+      "learning_rate": 2.1526856575547444e-06,
+      "loss": 0.6321,
+      "step": 686
+    },
+    {
+      "epoch": 2.1694736842105264,
+      "grad_norm": 0.547867476940155,
+      "learning_rate": 2.1375675893400373e-06,
+      "loss": 0.6192,
+      "step": 687
+    },
+    {
+      "epoch": 2.1726315789473682,
+      "grad_norm": 0.5950310826301575,
+      "learning_rate": 2.1224883483575166e-06,
+      "loss": 0.6292,
+      "step": 688
+    },
+    {
+      "epoch": 2.1757894736842105,
+      "grad_norm": 0.595939576625824,
+      "learning_rate": 2.1074481391483233e-06,
+      "loss": 0.6098,
+      "step": 689
+    },
+    {
+      "epoch": 2.1789473684210527,
+      "grad_norm": 0.6173756122589111,
+      "learning_rate": 2.0924471657241526e-06,
+      "loss": 0.6209,
+      "step": 690
+    },
+    {
+      "epoch": 2.1821052631578945,
+      "grad_norm": 0.5339811444282532,
+      "learning_rate": 2.0774856315644955e-06,
+      "loss": 0.6303,
+      "step": 691
+    },
+    {
+      "epoch": 2.185263157894737,
+      "grad_norm": 0.5910825133323669,
+      "learning_rate": 2.0625637396138666e-06,
+      "loss": 0.578,
+      "step": 692
+    },
+    {
+      "epoch": 2.188421052631579,
+      "grad_norm": 0.574038565158844,
+      "learning_rate": 2.0476816922790575e-06,
+      "loss": 0.5743,
+      "step": 693
+    },
+    {
+      "epoch": 2.191578947368421,
+      "grad_norm": 0.537571370601654,
+      "learning_rate": 2.0328396914263925e-06,
+      "loss": 0.6386,
+      "step": 694
+    },
+    {
+      "epoch": 2.194736842105263,
+      "grad_norm": 0.562436580657959,
+      "learning_rate": 2.0180379383789907e-06,
+      "loss": 0.6322,
+      "step": 695
+    },
+    {
+      "epoch": 2.1978947368421053,
+      "grad_norm": 0.5552737712860107,
+      "learning_rate": 2.0032766339140246e-06,
+      "loss": 0.6069,
+      "step": 696
+    },
+    {
+      "epoch": 2.201052631578947,
+      "grad_norm": 0.5092212557792664,
+      "learning_rate": 1.988555978260013e-06,
+      "loss": 0.6607,
+      "step": 697
+    },
+    {
+      "epoch": 2.2042105263157894,
+      "grad_norm": 0.558074414730072,
+      "learning_rate": 1.973876171094097e-06,
+      "loss": 0.6178,
+      "step": 698
+    },
+    {
+      "epoch": 2.2073684210526316,
+      "grad_norm": 0.5705953240394592,
+      "learning_rate": 1.9592374115393293e-06,
+      "loss": 0.6321,
+      "step": 699
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.5584784150123596,
+      "learning_rate": 1.9446398981619757e-06,
+      "loss": 0.5915,
+      "step": 700
+    },
+    {
+      "epoch": 2.2136842105263157,
+      "grad_norm": 0.5504412055015564,
+      "learning_rate": 1.9300838289688216e-06,
+      "loss": 0.6143,
+      "step": 701
+    },
+    {
+      "epoch": 2.216842105263158,
+      "grad_norm": 0.5174974203109741,
+      "learning_rate": 1.915569401404488e-06,
+      "loss": 0.5982,
+      "step": 702
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.5328599810600281,
+      "learning_rate": 1.9010968123487478e-06,
+      "loss": 0.6022,
+      "step": 703
+    },
+    {
+      "epoch": 2.223157894736842,
+      "grad_norm": 0.5421804785728455,
+      "learning_rate": 1.8866662581138646e-06,
+      "loss": 0.5817,
+      "step": 704
+    },
+    {
+      "epoch": 2.2263157894736842,
+      "grad_norm": 0.57915198802948,
+      "learning_rate": 1.8722779344419139e-06,
+      "loss": 0.613,
+      "step": 705
+    },
+    {
+      "epoch": 2.2294736842105265,
+      "grad_norm": 0.5778079628944397,
+      "learning_rate": 1.8579320365021508e-06,
+      "loss": 0.6301,
+      "step": 706
+    },
+    {
+      "epoch": 2.2326315789473683,
+      "grad_norm": 0.5267239212989807,
+      "learning_rate": 1.8436287588883416e-06,
+      "loss": 0.6269,
+      "step": 707
+    },
+    {
+      "epoch": 2.2357894736842105,
+      "grad_norm": 0.5386923551559448,
+      "learning_rate": 1.8293682956161357e-06,
+      "loss": 0.5903,
+      "step": 708
+    },
+    {
+      "epoch": 2.238947368421053,
+      "grad_norm": 0.5258237719535828,
+      "learning_rate": 1.8151508401204298e-06,
+      "loss": 0.6086,
+      "step": 709
+    },
+    {
+      "epoch": 2.2421052631578946,
+      "grad_norm": 0.5495336055755615,
+      "learning_rate": 1.800976585252745e-06,
+      "loss": 0.6247,
+      "step": 710
+    },
+    {
+      "epoch": 2.245263157894737,
+      "grad_norm": 0.55051589012146,
+      "learning_rate": 1.7868457232786117e-06,
+      "loss": 0.5962,
+      "step": 711
+    },
+    {
+      "epoch": 2.248421052631579,
+      "grad_norm": 0.5342631936073303,
+      "learning_rate": 1.7727584458749608e-06,
+      "loss": 0.587,
+      "step": 712
+    },
+    {
+      "epoch": 2.251578947368421,
+      "grad_norm": 0.5018291473388672,
+      "learning_rate": 1.7587149441275236e-06,
+      "loss": 0.6476,
+      "step": 713
+    },
+    {
+      "epoch": 2.254736842105263,
+      "grad_norm": 0.5503069758415222,
+      "learning_rate": 1.7447154085282398e-06,
+      "loss": 0.613,
+      "step": 714
+    },
+    {
+      "epoch": 2.2578947368421054,
+      "grad_norm": 0.5181307196617126,
+      "learning_rate": 1.7307600289726745e-06,
+      "loss": 0.6221,
+      "step": 715
+    },
+    {
+      "epoch": 2.261052631578947,
+      "grad_norm": 0.5215415358543396,
+      "learning_rate": 1.7168489947574407e-06,
+      "loss": 0.6311,
+      "step": 716
+    },
+    {
+      "epoch": 2.2642105263157895,
+      "grad_norm": 0.559654712677002,
+      "learning_rate": 1.7029824945776346e-06,
+      "loss": 0.5764,
+      "step": 717
+    },
+    {
+      "epoch": 2.2673684210526317,
+      "grad_norm": 0.5514314770698547,
+      "learning_rate": 1.6891607165242718e-06,
+      "loss": 0.629,
+      "step": 718
+    },
+    {
+      "epoch": 2.2705263157894735,
+      "grad_norm": 0.4893876910209656,
+      "learning_rate": 1.6753838480817397e-06,
+      "loss": 0.6328,
+      "step": 719
+    },
+    {
+      "epoch": 2.2736842105263158,
+      "grad_norm": 0.5276650190353394,
+      "learning_rate": 1.661652076125252e-06,
+      "loss": 0.5912,
+      "step": 720
+    },
+    {
+      "epoch": 2.276842105263158,
+      "grad_norm": 0.5329832434654236,
+      "learning_rate": 1.6479655869183142e-06,
+      "loss": 0.61,
+      "step": 721
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 0.535116970539093,
+      "learning_rate": 1.6343245661102031e-06,
+      "loss": 0.6346,
+      "step": 722
+    },
+    {
+      "epoch": 2.283157894736842,
+      "grad_norm": 0.5258433818817139,
+      "learning_rate": 1.620729198733434e-06,
+      "loss": 0.6336,
+      "step": 723
+    },
+    {
+      "epoch": 2.2863157894736843,
+      "grad_norm": 0.5541167259216309,
+      "learning_rate": 1.6071796692012663e-06,
+      "loss": 0.6221,
+      "step": 724
+    },
+    {
+      "epoch": 2.2894736842105265,
+      "grad_norm": 0.5115761160850525,
+      "learning_rate": 1.5936761613051937e-06,
+      "loss": 0.6489,
+      "step": 725
+    },
+    {
+      "epoch": 2.2926315789473684,
+      "grad_norm": 0.5598406195640564,
+      "learning_rate": 1.580218858212454e-06,
+      "loss": 0.6,
+      "step": 726
+    },
+    {
+      "epoch": 2.2957894736842106,
+      "grad_norm": 0.5273876190185547,
+      "learning_rate": 1.5668079424635424e-06,
+      "loss": 0.6227,
+      "step": 727
+    },
+    {
+      "epoch": 2.298947368421053,
+      "grad_norm": 0.5451186299324036,
+      "learning_rate": 1.5534435959697363e-06,
+      "loss": 0.5881,
+      "step": 728
+    },
+    {
+      "epoch": 2.3021052631578947,
+      "grad_norm": 0.5124930739402771,
+      "learning_rate": 1.5401260000106321e-06,
+      "loss": 0.6739,
+      "step": 729
+    },
+    {
+      "epoch": 2.305263157894737,
+      "grad_norm": 0.5419645309448242,
+      "learning_rate": 1.526855335231679e-06,
+      "loss": 0.6209,
+      "step": 730
+    },
+    {
+      "epoch": 2.308421052631579,
+      "grad_norm": 0.5517361760139465,
+      "learning_rate": 1.5136317816417333e-06,
+      "loss": 0.6345,
+      "step": 731
+    },
+    {
+      "epoch": 2.311578947368421,
+      "grad_norm": 0.5226176977157593,
+      "learning_rate": 1.5004555186106124e-06,
+      "loss": 0.6008,
+      "step": 732
+    },
+    {
+      "epoch": 2.314736842105263,
+      "grad_norm": 0.5727024078369141,
+      "learning_rate": 1.487326724866668e-06,
+      "loss": 0.5842,
+      "step": 733
+    },
+    {
+      "epoch": 2.3178947368421055,
+      "grad_norm": 0.5469027161598206,
+      "learning_rate": 1.4742455784943576e-06,
+      "loss": 0.6151,
+      "step": 734
+    },
+    {
+      "epoch": 2.3210526315789473,
+      "grad_norm": 0.5690147280693054,
+      "learning_rate": 1.4612122569318282e-06,
+      "loss": 0.566,
+      "step": 735
+    },
+    {
+      "epoch": 2.3242105263157895,
+      "grad_norm": 0.5304024815559387,
+      "learning_rate": 1.448226936968517e-06,
+      "loss": 0.6047,
+      "step": 736
+    },
+    {
+      "epoch": 2.3273684210526318,
+      "grad_norm": 0.5012169480323792,
+      "learning_rate": 1.4352897947427396e-06,
+      "loss": 0.6352,
+      "step": 737
+    },
+    {
+      "epoch": 2.3305263157894736,
+      "grad_norm": 0.49926042556762695,
+      "learning_rate": 1.422401005739314e-06,
+      "loss": 0.6673,
+      "step": 738
+    },
+    {
+      "epoch": 2.333684210526316,
+      "grad_norm": 0.5833967924118042,
+      "learning_rate": 1.4095607447871711e-06,
+      "loss": 0.5515,
+      "step": 739
+    },
+    {
+      "epoch": 2.336842105263158,
+      "grad_norm": 0.6136566996574402,
+      "learning_rate": 1.3967691860569915e-06,
+      "loss": 0.5947,
+      "step": 740
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.5746322274208069,
+      "learning_rate": 1.3840265030588323e-06,
+      "loss": 0.6308,
+      "step": 741
+    },
+    {
+      "epoch": 2.343157894736842,
+      "grad_norm": 0.5507466197013855,
+      "learning_rate": 1.3713328686397832e-06,
+      "loss": 0.6041,
+      "step": 742
+    },
+    {
+      "epoch": 2.3463157894736844,
+      "grad_norm": 0.524361252784729,
+      "learning_rate": 1.358688454981621e-06,
+      "loss": 0.5928,
+      "step": 743
+    },
+    {
+      "epoch": 2.349473684210526,
+      "grad_norm": 0.5451714396476746,
+      "learning_rate": 1.3460934335984677e-06,
+      "loss": 0.6086,
+      "step": 744
+    },
+    {
+      "epoch": 2.3526315789473684,
+      "grad_norm": 0.5549687147140503,
+      "learning_rate": 1.3335479753344688e-06,
+      "loss": 0.6241,
+      "step": 745
+    },
+    {
+      "epoch": 2.3557894736842107,
+      "grad_norm": 0.5424072742462158,
+      "learning_rate": 1.3210522503614753e-06,
+      "loss": 0.6078,
+      "step": 746
+    },
+    {
+      "epoch": 2.3589473684210525,
+      "grad_norm": 0.5681266188621521,
+      "learning_rate": 1.3086064281767346e-06,
+      "loss": 0.5994,
+      "step": 747
+    },
+    {
+      "epoch": 2.3621052631578947,
+      "grad_norm": 0.5240947008132935,
+      "learning_rate": 1.2962106776005917e-06,
+      "loss": 0.6148,
+      "step": 748
+    },
+    {
+      "epoch": 2.365263157894737,
+      "grad_norm": 0.5335183143615723,
+      "learning_rate": 1.2838651667742014e-06,
+      "loss": 0.6244,
+      "step": 749
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.5286924839019775,
+      "learning_rate": 1.2715700631572387e-06,
+      "loss": 0.6306,
+      "step": 750
+    },
+    {
+      "epoch": 2.371578947368421,
+      "grad_norm": 0.527306854724884,
+      "learning_rate": 1.2593255335256438e-06,
+      "loss": 0.64,
+      "step": 751
+    },
+    {
+      "epoch": 2.3747368421052633,
+      "grad_norm": 0.5724322199821472,
+      "learning_rate": 1.2471317439693436e-06,
+      "loss": 0.6075,
+      "step": 752
+    },
+    {
+      "epoch": 2.377894736842105,
+      "grad_norm": 0.5209251046180725,
+      "learning_rate": 1.2349888598900078e-06,
+      "loss": 0.5929,
+      "step": 753
+    },
+    {
+      "epoch": 2.3810526315789473,
+      "grad_norm": 0.53974848985672,
+      "learning_rate": 1.2228970459988015e-06,
+      "loss": 0.5841,
+      "step": 754
+    },
+    {
+      "epoch": 2.3842105263157896,
+      "grad_norm": 0.5406898260116577,
+      "learning_rate": 1.2108564663141541e-06,
+      "loss": 0.6004,
+      "step": 755
+    },
+    {
+      "epoch": 2.3873684210526314,
+      "grad_norm": 0.5027428269386292,
+      "learning_rate": 1.1988672841595312e-06,
+      "loss": 0.6057,
+      "step": 756
+    },
+    {
+      "epoch": 2.3905263157894736,
+      "grad_norm": 0.5120511054992676,
+      "learning_rate": 1.186929662161221e-06,
+      "loss": 0.6129,
+      "step": 757
+    },
+    {
+      "epoch": 2.393684210526316,
+      "grad_norm": 0.5876862406730652,
+      "learning_rate": 1.1750437622461293e-06,
+      "loss": 0.5984,
+      "step": 758
+    },
+    {
+      "epoch": 2.3968421052631577,
+      "grad_norm": 0.5318094491958618,
+      "learning_rate": 1.1632097456395802e-06,
+      "loss": 0.631,
+      "step": 759
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.48409977555274963,
+      "learning_rate": 1.1514277728631323e-06,
+      "loss": 0.6322,
+      "step": 760
+    },
+    {
+      "epoch": 2.403157894736842,
+      "grad_norm": 0.5341342687606812,
+      "learning_rate": 1.1396980037324e-06,
+      "loss": 0.6,
+      "step": 761
+    },
+    {
+      "epoch": 2.406315789473684,
+      "grad_norm": 0.5514808893203735,
+      "learning_rate": 1.128020597354884e-06,
+      "loss": 0.6348,
+      "step": 762
+    },
+    {
+      "epoch": 2.409473684210526,
+      "grad_norm": 0.5363562107086182,
+      "learning_rate": 1.1163957121278163e-06,
+      "loss": 0.5762,
+      "step": 763
+    },
+    {
+      "epoch": 2.4126315789473685,
+      "grad_norm": 0.5589635968208313,
+      "learning_rate": 1.104823505736009e-06,
+      "loss": 0.6406,
+      "step": 764
+    },
+    {
+      "epoch": 2.4157894736842107,
+      "grad_norm": 0.5272628664970398,
+      "learning_rate": 1.093304135149717e-06,
+      "loss": 0.5923,
+      "step": 765
+    },
+    {
+      "epoch": 2.4189473684210525,
+      "grad_norm": 0.54401695728302,
+      "learning_rate": 1.0818377566225075e-06,
+      "loss": 0.5936,
+      "step": 766
+    },
+    {
+      "epoch": 2.4221052631578948,
+      "grad_norm": 0.5256767272949219,
+      "learning_rate": 1.070424525689142e-06,
+      "loss": 0.6173,
+      "step": 767
+    },
+    {
+      "epoch": 2.425263157894737,
+      "grad_norm": 0.5305107831954956,
+      "learning_rate": 1.0590645971634655e-06,
+      "loss": 0.5995,
+      "step": 768
+    },
+    {
+      "epoch": 2.428421052631579,
+      "grad_norm": 0.5133773684501648,
+      "learning_rate": 1.0477581251363066e-06,
+      "loss": 0.6294,
+      "step": 769
+    },
+    {
+      "epoch": 2.431578947368421,
+      "grad_norm": 0.5675320625305176,
+      "learning_rate": 1.0365052629733884e-06,
+      "loss": 0.5941,
+      "step": 770
+    },
+    {
+      "epoch": 2.4347368421052633,
+      "grad_norm": 0.5107877254486084,
+      "learning_rate": 1.025306163313246e-06,
+      "loss": 0.6195,
+      "step": 771
+    },
+    {
+      "epoch": 2.437894736842105,
+      "grad_norm": 0.5630270838737488,
+      "learning_rate": 1.0141609780651585e-06,
+      "loss": 0.6353,
+      "step": 772
+    },
+    {
+      "epoch": 2.4410526315789474,
+      "grad_norm": 0.4963136911392212,
+      "learning_rate": 1.0030698584070848e-06,
+      "loss": 0.6019,
+      "step": 773
+    },
+    {
+      "epoch": 2.4442105263157896,
+      "grad_norm": 0.5430152416229248,
+      "learning_rate": 9.92032954783621e-07,
+      "loss": 0.6518,
+      "step": 774
+    },
+    {
+      "epoch": 2.4473684210526314,
+      "grad_norm": 0.5275354981422424,
+      "learning_rate": 9.81050416903951e-07,
+      "loss": 0.6547,
+      "step": 775
+    },
+    {
+      "epoch": 2.4505263157894737,
+      "grad_norm": 0.5483253598213196,
+      "learning_rate": 9.701223937398152e-07,
+      "loss": 0.634,
+      "step": 776
+    },
+    {
+      "epoch": 2.453684210526316,
+      "grad_norm": 0.5382000803947449,
+      "learning_rate": 9.592490335234993e-07,
+      "loss": 0.648,
+      "step": 777
+    },
+    {
+      "epoch": 2.4568421052631577,
+      "grad_norm": 0.5335687398910522,
+      "learning_rate": 9.484304837458158e-07,
+      "loss": 0.5985,
+      "step": 778
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.5186282992362976,
+      "learning_rate": 9.376668911541042e-07,
+      "loss": 0.6092,
+      "step": 779
+    },
+    {
+      "epoch": 2.463157894736842,
+      "grad_norm": 0.517254114151001,
+      "learning_rate": 9.269584017502431e-07,
+      "loss": 0.6524,
+      "step": 780
+    },
+    {
+      "epoch": 2.466315789473684,
+      "grad_norm": 0.5538238286972046,
+      "learning_rate": 9.163051607886703e-07,
+      "loss": 0.6008,
+      "step": 781
+    },
+    {
+      "epoch": 2.4694736842105263,
+      "grad_norm": 0.5268839597702026,
+      "learning_rate": 9.057073127744065e-07,
+      "loss": 0.6115,
+      "step": 782
+    },
+    {
+      "epoch": 2.4726315789473685,
+      "grad_norm": 0.5400952696800232,
+      "learning_rate": 8.951650014611019e-07,
+      "loss": 0.5816,
+      "step": 783
+    },
+    {
+      "epoch": 2.4757894736842108,
+      "grad_norm": 0.571423351764679,
+      "learning_rate": 8.846783698490835e-07,
+      "loss": 0.6134,
+      "step": 784
+    },
+    {
+      "epoch": 2.4789473684210526,
+      "grad_norm": 0.5169550180435181,
+      "learning_rate": 8.742475601834133e-07,
+      "loss": 0.6033,
+      "step": 785
+    },
+    {
+      "epoch": 2.482105263157895,
+      "grad_norm": 0.5374566912651062,
+      "learning_rate": 8.638727139519637e-07,
+      "loss": 0.597,
+      "step": 786
+    },
+    {
+      "epoch": 2.485263157894737,
+      "grad_norm": 0.5302397012710571,
+      "learning_rate": 8.535539718834929e-07,
+      "loss": 0.6735,
+      "step": 787
+    },
+    {
+      "epoch": 2.488421052631579,
+      "grad_norm": 0.5013649463653564,
+      "learning_rate": 8.432914739457432e-07,
+      "loss": 0.6714,
+      "step": 788
+    },
+    {
+      "epoch": 2.491578947368421,
+      "grad_norm": 0.5414134860038757,
+      "learning_rate": 8.330853593435345e-07,
+      "loss": 0.6105,
+      "step": 789
+    },
+    {
+      "epoch": 2.4947368421052634,
+      "grad_norm": 0.5048105120658875,
+      "learning_rate": 8.229357665168791e-07,
+      "loss": 0.6186,
+      "step": 790
+    },
+    {
+      "epoch": 2.497894736842105,
+      "grad_norm": 0.5406049489974976,
+      "learning_rate": 8.12842833139107e-07,
+      "loss": 0.6146,
+      "step": 791
+    },
+    {
+      "epoch": 2.5010526315789474,
+      "grad_norm": 0.5164625644683838,
+      "learning_rate": 8.028066961149921e-07,
+      "loss": 0.6194,
+      "step": 792
+    },
+    {
+      "epoch": 2.5042105263157897,
+      "grad_norm": 0.5256004929542542,
+      "learning_rate": 7.928274915789035e-07,
+      "loss": 0.6136,
+      "step": 793
+    },
+    {
+      "epoch": 2.5073684210526315,
+      "grad_norm": 0.5570123791694641,
+      "learning_rate": 7.829053548929488e-07,
+      "loss": 0.6325,
+      "step": 794
+    },
+    {
+      "epoch": 2.5105263157894737,
+      "grad_norm": 0.5029011964797974,
+      "learning_rate": 7.730404206451459e-07,
+      "loss": 0.599,
+      "step": 795
+    },
+    {
+      "epoch": 2.513684210526316,
+      "grad_norm": 0.5093344449996948,
+      "learning_rate": 7.632328226475971e-07,
+      "loss": 0.6007,
+      "step": 796
+    },
+    {
+      "epoch": 2.5168421052631578,
+      "grad_norm": 0.4984530508518219,
+      "learning_rate": 7.53482693934669e-07,
+      "loss": 0.639,
+      "step": 797
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.5288375616073608,
+      "learning_rate": 7.437901667611908e-07,
+      "loss": 0.6021,
+      "step": 798
+    },
+    {
+      "epoch": 2.5231578947368423,
+      "grad_norm": 0.5130994915962219,
+      "learning_rate": 7.341553726006611e-07,
+      "loss": 0.6125,
+      "step": 799
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.5277830958366394,
+      "learning_rate": 7.245784421434643e-07,
+      "loss": 0.6388,
+      "step": 800
+    },
+    {
+      "epoch": 2.5294736842105263,
+      "grad_norm": 0.5155095458030701,
+      "learning_rate": 7.150595052950954e-07,
+      "loss": 0.6319,
+      "step": 801
+    },
+    {
+      "epoch": 2.5326315789473686,
+      "grad_norm": 0.5809716582298279,
+      "learning_rate": 7.055986911744017e-07,
+      "loss": 0.578,
+      "step": 802
+    },
+    {
+      "epoch": 2.5357894736842104,
+      "grad_norm": 0.5170502066612244,
+      "learning_rate": 6.961961281118285e-07,
+      "loss": 0.6074,
+      "step": 803
+    },
+    {
+      "epoch": 2.5389473684210526,
+      "grad_norm": 0.5173423290252686,
+      "learning_rate": 6.868519436476795e-07,
+      "loss": 0.631,
+      "step": 804
+    },
+    {
+      "epoch": 2.542105263157895,
+      "grad_norm": 0.49472615122795105,
+      "learning_rate": 6.775662645303871e-07,
+      "loss": 0.6439,
+      "step": 805
+    },
+    {
+      "epoch": 2.5452631578947367,
+      "grad_norm": 0.5428845882415771,
+      "learning_rate": 6.683392167147917e-07,
+      "loss": 0.6203,
+      "step": 806
+    },
+    {
+      "epoch": 2.548421052631579,
+      "grad_norm": 0.5209921002388,
+      "learning_rate": 6.591709253604356e-07,
+      "loss": 0.6588,
+      "step": 807
+    },
+    {
+      "epoch": 2.551578947368421,
+      "grad_norm": 0.5301423668861389,
+      "learning_rate": 6.500615148298617e-07,
+      "loss": 0.6003,
+      "step": 808
+    },
+    {
+      "epoch": 2.554736842105263,
+      "grad_norm": 0.5289193391799927,
+      "learning_rate": 6.410111086869314e-07,
+      "loss": 0.6389,
+      "step": 809
+    },
+    {
+      "epoch": 2.557894736842105,
+      "grad_norm": 0.5304815769195557,
+      "learning_rate": 6.320198296951435e-07,
+      "loss": 0.5732,
+      "step": 810
+    },
+    {
+      "epoch": 2.5610526315789475,
+      "grad_norm": 0.5228027701377869,
+      "learning_rate": 6.230877998159724e-07,
+      "loss": 0.6278,
+      "step": 811
+    },
+    {
+      "epoch": 2.5642105263157893,
+      "grad_norm": 0.5397827625274658,
+      "learning_rate": 6.142151402072133e-07,
+      "loss": 0.5979,
+      "step": 812
+    },
+    {
+      "epoch": 2.5673684210526315,
+      "grad_norm": 0.5186311602592468,
+      "learning_rate": 6.054019712213377e-07,
+      "loss": 0.5872,
+      "step": 813
+    },
+    {
+      "epoch": 2.5705263157894738,
+      "grad_norm": 0.5260378122329712,
+      "learning_rate": 5.966484124038602e-07,
+      "loss": 0.6337,
+      "step": 814
+    },
+    {
+      "epoch": 2.5736842105263156,
+      "grad_norm": 0.556025505065918,
+      "learning_rate": 5.879545824917199e-07,
+      "loss": 0.5965,
+      "step": 815
+    },
+    {
+      "epoch": 2.576842105263158,
+      "grad_norm": 0.5116238594055176,
+      "learning_rate": 5.793205994116674e-07,
+      "loss": 0.6147,
+      "step": 816
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.5269439816474915,
+      "learning_rate": 5.707465802786655e-07,
+      "loss": 0.5967,
+      "step": 817
+    },
+    {
+      "epoch": 2.583157894736842,
+      "grad_norm": 0.5351024270057678,
+      "learning_rate": 5.622326413942997e-07,
+      "loss": 0.6241,
+      "step": 818
+    },
+    {
+      "epoch": 2.586315789473684,
+      "grad_norm": 0.536216676235199,
+      "learning_rate": 5.537788982452052e-07,
+      "loss": 0.631,
+      "step": 819
+    },
+    {
+      "epoch": 2.5894736842105264,
+      "grad_norm": 0.5194189548492432,
+      "learning_rate": 5.453854655014956e-07,
+      "loss": 0.6394,
+      "step": 820
+    },
+    {
+      "epoch": 2.592631578947368,
+      "grad_norm": 0.5215575098991394,
+      "learning_rate": 5.370524570152059e-07,
+      "loss": 0.5948,
+      "step": 821
+    },
+    {
+      "epoch": 2.5957894736842104,
+      "grad_norm": 0.5416327118873596,
+      "learning_rate": 5.287799858187548e-07,
+      "loss": 0.6352,
+      "step": 822
+    },
+    {
+      "epoch": 2.5989473684210527,
+      "grad_norm": 0.5264502167701721,
+      "learning_rate": 5.205681641234062e-07,
+      "loss": 0.61,
+      "step": 823
+    },
+    {
+      "epoch": 2.6021052631578945,
+      "grad_norm": 0.5675118565559387,
+      "learning_rate": 5.1241710331775e-07,
+      "loss": 0.5791,
+      "step": 824
+    },
+    {
+      "epoch": 2.6052631578947367,
+      "grad_norm": 0.49685797095298767,
+      "learning_rate": 5.043269139661872e-07,
+      "loss": 0.6316,
+      "step": 825
+    },
+    {
+      "epoch": 2.608421052631579,
+      "grad_norm": 0.5199131965637207,
+      "learning_rate": 4.962977058074381e-07,
+      "loss": 0.6217,
+      "step": 826
+    },
+    {
+      "epoch": 2.6115789473684208,
+      "grad_norm": 0.5157626271247864,
+      "learning_rate": 4.883295877530431e-07,
+      "loss": 0.6394,
+      "step": 827
+    },
+    {
+      "epoch": 2.614736842105263,
+      "grad_norm": 0.5056947469711304,
+      "learning_rate": 4.804226678858936e-07,
+      "loss": 0.6168,
+      "step": 828
+    },
+    {
+      "epoch": 2.6178947368421053,
+      "grad_norm": 0.5152086019515991,
+      "learning_rate": 4.725770534587637e-07,
+      "loss": 0.5963,
+      "step": 829
+    },
+    {
+      "epoch": 2.6210526315789475,
+      "grad_norm": 0.4946594536304474,
+      "learning_rate": 4.647928508928512e-07,
+      "loss": 0.6344,
+      "step": 830
+    },
+    {
+      "epoch": 2.6242105263157893,
+      "grad_norm": 0.5005388855934143,
+      "learning_rate": 4.5707016577634156e-07,
+      "loss": 0.6797,
+      "step": 831
+    },
+    {
+      "epoch": 2.6273684210526316,
+      "grad_norm": 0.49225690960884094,
+      "learning_rate": 4.494091028629699e-07,
+      "loss": 0.6552,
+      "step": 832
+    },
+    {
+      "epoch": 2.630526315789474,
+      "grad_norm": 0.5395132303237915,
+      "learning_rate": 4.418097660706039e-07,
+      "loss": 0.6197,
+      "step": 833
+    },
+    {
+      "epoch": 2.6336842105263156,
+      "grad_norm": 0.5258274674415588,
+      "learning_rate": 4.342722584798298e-07,
+      "loss": 0.6151,
+      "step": 834
+    },
+    {
+      "epoch": 2.636842105263158,
+      "grad_norm": 0.507228672504425,
+      "learning_rate": 4.267966823325581e-07,
+      "loss": 0.6213,
+      "step": 835
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.5385991334915161,
+      "learning_rate": 4.193831390306352e-07,
+      "loss": 0.6423,
+      "step": 836
+    },
+    {
+      "epoch": 2.6431578947368424,
+      "grad_norm": 0.5073803067207336,
+      "learning_rate": 4.1203172913446774e-07,
+      "loss": 0.5812,
+      "step": 837
+    },
+    {
+      "epoch": 2.646315789473684,
+      "grad_norm": 0.5264935493469238,
+      "learning_rate": 4.047425523616577e-07,
+      "loss": 0.6305,
+      "step": 838
+    },
+    {
+      "epoch": 2.6494736842105264,
+      "grad_norm": 0.4937891364097595,
+      "learning_rate": 3.9751570758565284e-07,
+      "loss": 0.6197,
+      "step": 839
+    },
+    {
+      "epoch": 2.6526315789473687,
+      "grad_norm": 0.5468772053718567,
+      "learning_rate": 3.9035129283440165e-07,
+      "loss": 0.592,
+      "step": 840
+    },
+    {
+      "epoch": 2.6557894736842105,
+      "grad_norm": 0.5518105626106262,
+      "learning_rate": 3.8324940528902845e-07,
+      "loss": 0.6332,
+      "step": 841
+    },
+    {
+      "epoch": 2.6589473684210527,
+      "grad_norm": 0.5344856381416321,
+      "learning_rate": 3.762101412825098e-07,
+      "loss": 0.6221,
+      "step": 842
+    },
+    {
+      "epoch": 2.662105263157895,
+      "grad_norm": 0.527391791343689,
+      "learning_rate": 3.6923359629837117e-07,
+      "loss": 0.5941,
+      "step": 843
+    },
+    {
+      "epoch": 2.665263157894737,
+      "grad_norm": 0.5052513480186462,
+      "learning_rate": 3.6231986496939153e-07,
+      "loss": 0.6352,
+      "step": 844
+    },
+    {
+      "epoch": 2.668421052631579,
+      "grad_norm": 0.5419393181800842,
+      "learning_rate": 3.554690410763173e-07,
+      "loss": 0.6058,
+      "step": 845
+    },
+    {
+      "epoch": 2.6715789473684213,
+      "grad_norm": 0.5119031667709351,
+      "learning_rate": 3.4868121754659533e-07,
+      "loss": 0.5942,
+      "step": 846
+    },
+    {
+      "epoch": 2.674736842105263,
+      "grad_norm": 0.4917730689048767,
+      "learning_rate": 3.4195648645310443e-07,
+      "loss": 0.6145,
+      "step": 847
+    },
+    {
+      "epoch": 2.6778947368421053,
+      "grad_norm": 0.5071961283683777,
+      "learning_rate": 3.3529493901291567e-07,
+      "loss": 0.6481,
+      "step": 848
+    },
+    {
+      "epoch": 2.6810526315789476,
+      "grad_norm": 0.5443476438522339,
+      "learning_rate": 3.286966655860485e-07,
+      "loss": 0.5775,
+      "step": 849
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.5274863839149475,
+      "learning_rate": 3.2216175567424737e-07,
+      "loss": 0.5835,
+      "step": 850
+    },
+    {
+      "epoch": 2.6873684210526316,
+      "grad_norm": 0.5420557856559753,
+      "learning_rate": 3.156902979197679e-07,
+      "loss": 0.5913,
+      "step": 851
+    },
+    {
+      "epoch": 2.690526315789474,
+      "grad_norm": 0.5218223929405212,
+      "learning_rate": 3.0928238010417275e-07,
+      "loss": 0.6144,
+      "step": 852
+    },
+    {
+      "epoch": 2.6936842105263157,
+      "grad_norm": 0.5233944654464722,
+      "learning_rate": 3.029380891471445e-07,
+      "loss": 0.632,
+      "step": 853
+    },
+    {
+      "epoch": 2.696842105263158,
+      "grad_norm": 0.5044023394584656,
+      "learning_rate": 2.966575111053027e-07,
+      "loss": 0.5874,
+      "step": 854
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.5117988586425781,
+      "learning_rate": 2.9044073117103777e-07,
+      "loss": 0.6533,
+      "step": 855
+    },
+    {
+      "epoch": 2.703157894736842,
+      "grad_norm": 0.511157751083374,
+      "learning_rate": 2.842878336713578e-07,
+      "loss": 0.6282,
+      "step": 856
+    },
+    {
+      "epoch": 2.7063157894736842,
+      "grad_norm": 0.5114948749542236,
+      "learning_rate": 2.7819890206674083e-07,
+      "loss": 0.6355,
+      "step": 857
+    },
+    {
+      "epoch": 2.7094736842105265,
+      "grad_norm": 0.549742579460144,
+      "learning_rate": 2.7217401895000664e-07,
+      "loss": 0.6086,
+      "step": 858
+    },
+    {
+      "epoch": 2.7126315789473683,
+      "grad_norm": 0.5112407803535461,
+      "learning_rate": 2.6621326604519216e-07,
+      "loss": 0.6062,
+      "step": 859
+    },
+    {
+      "epoch": 2.7157894736842105,
+      "grad_norm": 0.5380560755729675,
+      "learning_rate": 2.6031672420644694e-07,
+      "loss": 0.5904,
+      "step": 860
+    },
+    {
+      "epoch": 2.718947368421053,
+      "grad_norm": 0.49735158681869507,
+      "learning_rate": 2.5448447341693493e-07,
+      "loss": 0.593,
+      "step": 861
+    },
+    {
+      "epoch": 2.7221052631578946,
+      "grad_norm": 0.4785706698894501,
+      "learning_rate": 2.4871659278774884e-07,
+      "loss": 0.6211,
+      "step": 862
+    },
+    {
+      "epoch": 2.725263157894737,
+      "grad_norm": 0.5243927240371704,
+      "learning_rate": 2.430131605568353e-07,
+      "loss": 0.6143,
+      "step": 863
+    },
+    {
+      "epoch": 2.728421052631579,
+      "grad_norm": 0.5355093479156494,
+      "learning_rate": 2.3737425408794202e-07,
+      "loss": 0.6162,
+      "step": 864
+    },
+    {
+      "epoch": 2.731578947368421,
+      "grad_norm": 0.5540766716003418,
+      "learning_rate": 2.31799949869555e-07,
+      "loss": 0.5777,
+      "step": 865
+    },
+    {
+      "epoch": 2.734736842105263,
+      "grad_norm": 0.5112879276275635,
+      "learning_rate": 2.2629032351387247e-07,
+      "loss": 0.5949,
+      "step": 866
+    },
+    {
+      "epoch": 2.7378947368421054,
+      "grad_norm": 0.5168077945709229,
+      "learning_rate": 2.2084544975577383e-07,
+      "loss": 0.6391,
+      "step": 867
+    },
+    {
+      "epoch": 2.741052631578947,
+      "grad_norm": 0.5143442749977112,
+      "learning_rate": 2.1546540245180825e-07,
+      "loss": 0.6142,
+      "step": 868
+    },
+    {
+      "epoch": 2.7442105263157894,
+      "grad_norm": 0.49836885929107666,
+      "learning_rate": 2.1015025457919002e-07,
+      "loss": 0.5926,
+      "step": 869
+    },
+    {
+      "epoch": 2.7473684210526317,
+      "grad_norm": 0.5337677597999573,
+      "learning_rate": 2.0490007823481096e-07,
+      "loss": 0.5758,
+      "step": 870
+    },
+    {
+      "epoch": 2.7505263157894735,
+      "grad_norm": 0.5162203311920166,
+      "learning_rate": 1.9971494463426332e-07,
+      "loss": 0.6044,
+      "step": 871
+    },
+    {
+      "epoch": 2.7536842105263157,
+      "grad_norm": 0.5083566308021545,
+      "learning_rate": 1.9459492411087078e-07,
+      "loss": 0.6377,
+      "step": 872
+    },
+    {
+      "epoch": 2.756842105263158,
+      "grad_norm": 0.5102013349533081,
+      "learning_rate": 1.8954008611473618e-07,
+      "loss": 0.6287,
+      "step": 873
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.5197393894195557,
+      "learning_rate": 1.8455049921179858e-07,
+      "loss": 0.6224,
+      "step": 874
+    },
+    {
+      "epoch": 2.763157894736842,
+      "grad_norm": 0.4867171347141266,
+      "learning_rate": 1.7962623108290556e-07,
+      "loss": 0.6475,
+      "step": 875
+    },
+    {
+      "epoch": 2.7663157894736843,
+      "grad_norm": 0.5163273215293884,
+      "learning_rate": 1.7476734852289235e-07,
+      "loss": 0.6408,
+      "step": 876
+    },
+    {
+      "epoch": 2.769473684210526,
+      "grad_norm": 0.5355550646781921,
+      "learning_rate": 1.6997391743967696e-07,
+      "loss": 0.6129,
+      "step": 877
+    },
+    {
+      "epoch": 2.7726315789473683,
+      "grad_norm": 0.5214855670928955,
+      "learning_rate": 1.65246002853367e-07,
+      "loss": 0.5987,
+      "step": 878
+    },
+    {
+      "epoch": 2.7757894736842106,
+      "grad_norm": 0.5373660922050476,
+      "learning_rate": 1.6058366889537546e-07,
+      "loss": 0.5472,
+      "step": 879
+    },
+    {
+      "epoch": 2.7789473684210524,
+      "grad_norm": 0.49646681547164917,
+      "learning_rate": 1.559869788075541e-07,
+      "loss": 0.612,
+      "step": 880
+    },
+    {
+      "epoch": 2.7821052631578946,
+      "grad_norm": 0.5255357623100281,
+      "learning_rate": 1.514559949413319e-07,
+      "loss": 0.6083,
+      "step": 881
+    },
+    {
+      "epoch": 2.785263157894737,
+      "grad_norm": 0.5195873379707336,
+      "learning_rate": 1.4699077875687252e-07,
+      "loss": 0.6032,
+      "step": 882
+    },
+    {
+      "epoch": 2.7884210526315787,
+      "grad_norm": 0.5284191370010376,
+      "learning_rate": 1.4259139082223761e-07,
+      "loss": 0.6466,
+      "step": 883
+    },
+    {
+      "epoch": 2.791578947368421,
+      "grad_norm": 0.5327157378196716,
+      "learning_rate": 1.3825789081256812e-07,
+      "loss": 0.5883,
+      "step": 884
+    },
+    {
+      "epoch": 2.794736842105263,
+      "grad_norm": 0.5047375559806824,
+      "learning_rate": 1.3399033750927327e-07,
+      "loss": 0.6168,
+      "step": 885
+    },
+    {
+      "epoch": 2.797894736842105,
+      "grad_norm": 0.5161851644515991,
+      "learning_rate": 1.297887887992344e-07,
+      "loss": 0.615,
+      "step": 886
+    },
+    {
+      "epoch": 2.8010526315789472,
+      "grad_norm": 0.5051410794258118,
+      "learning_rate": 1.2565330167401747e-07,
+      "loss": 0.6444,
+      "step": 887
+    },
+    {
+      "epoch": 2.8042105263157895,
+      "grad_norm": 0.5022399425506592,
+      "learning_rate": 1.2158393222910235e-07,
+      "loss": 0.6318,
+      "step": 888
+    },
+    {
+      "epoch": 2.8073684210526317,
+      "grad_norm": 0.4974055588245392,
+      "learning_rate": 1.175807356631209e-07,
+      "loss": 0.6291,
+      "step": 889
+    },
+    {
+      "epoch": 2.8105263157894735,
+      "grad_norm": 0.4854937493801117,
+      "learning_rate": 1.1364376627710727e-07,
+      "loss": 0.6456,
+      "step": 890
+    },
+    {
+      "epoch": 2.813684210526316,
+      "grad_norm": 0.5069525837898254,
+      "learning_rate": 1.0977307747376431e-07,
+      "loss": 0.6226,
+      "step": 891
+    },
+    {
+      "epoch": 2.816842105263158,
+      "grad_norm": 0.5431550145149231,
+      "learning_rate": 1.0596872175673456e-07,
+      "loss": 0.6005,
+      "step": 892
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.5111134052276611,
+      "learning_rate": 1.0223075072989418e-07,
+      "loss": 0.6243,
+      "step": 893
+    },
+    {
+      "epoch": 2.823157894736842,
+      "grad_norm": 0.5054842829704285,
+      "learning_rate": 9.855921509664745e-08,
+      "loss": 0.6145,
+      "step": 894
+    },
+    {
+      "epoch": 2.8263157894736843,
+      "grad_norm": 0.5049132108688354,
+      "learning_rate": 9.495416465924113e-08,
+      "loss": 0.6283,
+      "step": 895
+    },
+    {
+      "epoch": 2.8294736842105266,
+      "grad_norm": 0.5091620087623596,
+      "learning_rate": 9.141564831808947e-08,
+      "loss": 0.6245,
+      "step": 896
+    },
+    {
+      "epoch": 2.8326315789473684,
+      "grad_norm": 0.5132347941398621,
+      "learning_rate": 8.794371407111091e-08,
+      "loss": 0.6149,
+      "step": 897
+    },
+    {
+      "epoch": 2.8357894736842106,
+      "grad_norm": 0.5056635737419128,
+      "learning_rate": 8.45384090130752e-08,
+      "loss": 0.6408,
+      "step": 898
+    },
+    {
+      "epoch": 2.838947368421053,
+      "grad_norm": 0.4924507141113281,
+      "learning_rate": 8.11997793349667e-08,
+      "loss": 0.622,
+      "step": 899
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.4771850109100342,
+      "learning_rate": 7.792787032335657e-08,
+      "loss": 0.6321,
+      "step": 900
+    },
+    {
+      "epoch": 2.845263157894737,
+      "grad_norm": 0.5036438703536987,
+      "learning_rate": 7.472272635978995e-08,
+      "loss": 0.6403,
+      "step": 901
+    },
+    {
+      "epoch": 2.848421052631579,
+      "grad_norm": 0.5470741987228394,
+      "learning_rate": 7.158439092018077e-08,
+      "loss": 0.6147,
+      "step": 902
+    },
+    {
+      "epoch": 2.851578947368421,
+      "grad_norm": 0.4736262261867523,
+      "learning_rate": 6.851290657422627e-08,
+      "loss": 0.6591,
+      "step": 903
+    },
+    {
+      "epoch": 2.8547368421052632,
+      "grad_norm": 0.4981045424938202,
+      "learning_rate": 6.550831498482679e-08,
+      "loss": 0.5957,
+      "step": 904
+    },
+    {
+      "epoch": 2.8578947368421055,
+      "grad_norm": 0.5055891275405884,
+      "learning_rate": 6.257065690752129e-08,
+      "loss": 0.6222,
+      "step": 905
+    },
+    {
+      "epoch": 2.8610526315789473,
+      "grad_norm": 0.5165517330169678,
+      "learning_rate": 5.969997218993328e-08,
+      "loss": 0.6335,
+      "step": 906
+    },
+    {
+      "epoch": 2.8642105263157895,
+      "grad_norm": 0.5130405426025391,
+      "learning_rate": 5.689629977123412e-08,
+      "loss": 0.602,
+      "step": 907
+    },
+    {
+      "epoch": 2.867368421052632,
+      "grad_norm": 0.5092564821243286,
+      "learning_rate": 5.415967768160946e-08,
+      "loss": 0.6221,
+      "step": 908
+    },
+    {
+      "epoch": 2.8705263157894736,
+      "grad_norm": 0.5033932328224182,
+      "learning_rate": 5.149014304174915e-08,
+      "loss": 0.632,
+      "step": 909
+    },
+    {
+      "epoch": 2.873684210526316,
+      "grad_norm": 0.5016679167747498,
+      "learning_rate": 4.8887732062337656e-08,
+      "loss": 0.61,
+      "step": 910
+    },
+    {
+      "epoch": 2.876842105263158,
+      "grad_norm": 0.5133535265922546,
+      "learning_rate": 4.635248004356885e-08,
+      "loss": 0.6702,
+      "step": 911
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.496550977230072,
+      "learning_rate": 4.388442137466198e-08,
+      "loss": 0.6243,
+      "step": 912
+    },
+    {
+      "epoch": 2.883157894736842,
+      "grad_norm": 0.48129701614379883,
+      "learning_rate": 4.148358953339926e-08,
+      "loss": 0.6565,
+      "step": 913
+    },
+    {
+      "epoch": 2.8863157894736844,
+      "grad_norm": 0.5186068415641785,
+      "learning_rate": 3.9150017085669566e-08,
+      "loss": 0.63,
+      "step": 914
+    },
+    {
+      "epoch": 2.889473684210526,
+      "grad_norm": 0.5132560133934021,
+      "learning_rate": 3.688373568502601e-08,
+      "loss": 0.6137,
+      "step": 915
+    },
+    {
+      "epoch": 2.8926315789473684,
+      "grad_norm": 0.5437251329421997,
+      "learning_rate": 3.468477607226017e-08,
+      "loss": 0.6646,
+      "step": 916
+    },
+    {
+      "epoch": 2.8957894736842107,
+      "grad_norm": 0.5179529786109924,
+      "learning_rate": 3.255316807498077e-08,
+      "loss": 0.5848,
+      "step": 917
+    },
+    {
+      "epoch": 2.8989473684210525,
+      "grad_norm": 0.5003638863563538,
+      "learning_rate": 3.048894060721064e-08,
+      "loss": 0.6227,
+      "step": 918
+    },
+    {
+      "epoch": 2.9021052631578947,
+      "grad_norm": 0.5409918427467346,
+      "learning_rate": 2.8492121668997064e-08,
+      "loss": 0.6075,
+      "step": 919
+    },
+    {
+      "epoch": 2.905263157894737,
+      "grad_norm": 0.4994703531265259,
+      "learning_rate": 2.6562738346027627e-08,
+      "loss": 0.6214,
+      "step": 920
+    },
+    {
+      "epoch": 2.908421052631579,
+      "grad_norm": 0.5022985935211182,
+      "learning_rate": 2.4700816809266615e-08,
+      "loss": 0.6123,
+      "step": 921
+    },
+    {
+      "epoch": 2.911578947368421,
+      "grad_norm": 0.5189907550811768,
+      "learning_rate": 2.290638231459641e-08,
+      "loss": 0.5687,
+      "step": 922
+    },
+    {
+      "epoch": 2.9147368421052633,
+      "grad_norm": 0.4975310266017914,
+      "learning_rate": 2.1179459202479436e-08,
+      "loss": 0.6152,
+      "step": 923
+    },
+    {
+      "epoch": 2.917894736842105,
+      "grad_norm": 0.5323106050491333,
+      "learning_rate": 1.9520070897623976e-08,
+      "loss": 0.6638,
+      "step": 924
+    },
+    {
+      "epoch": 2.9210526315789473,
+      "grad_norm": 0.48443955183029175,
+      "learning_rate": 1.792823990866721e-08,
+      "loss": 0.6238,
+      "step": 925
+    },
+    {
+      "epoch": 2.9242105263157896,
+      "grad_norm": 0.48831799626350403,
+      "learning_rate": 1.640398782787267e-08,
+      "loss": 0.5758,
+      "step": 926
+    },
+    {
+      "epoch": 2.9273684210526314,
+      "grad_norm": 0.49946728348731995,
+      "learning_rate": 1.49473353308327e-08,
+      "loss": 0.6192,
+      "step": 927
+    },
+    {
+      "epoch": 2.9305263157894736,
+      "grad_norm": 0.5276715159416199,
+      "learning_rate": 1.3558302176192584e-08,
+      "loss": 0.6171,
+      "step": 928
+    },
+    {
+      "epoch": 2.933684210526316,
+      "grad_norm": 0.5295907855033875,
+      "learning_rate": 1.2236907205379623e-08,
+      "loss": 0.6198,
+      "step": 929
+    },
+    {
+      "epoch": 2.9368421052631577,
+      "grad_norm": 0.4946899712085724,
+      "learning_rate": 1.0983168342348915e-08,
+      "loss": 0.6253,
+      "step": 930
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.5253959894180298,
+      "learning_rate": 9.797102593339659e-09,
+      "loss": 0.5993,
+      "step": 931
+    },
+    {
+      "epoch": 2.943157894736842,
+      "grad_norm": 0.5255020260810852,
+      "learning_rate": 8.678726046644215e-09,
+      "loss": 0.6384,
+      "step": 932
+    },
+    {
+      "epoch": 2.946315789473684,
+      "grad_norm": 0.5470498204231262,
+      "learning_rate": 7.628053872390517e-09,
+      "loss": 0.6002,
+      "step": 933
+    },
+    {
+      "epoch": 2.9494736842105262,
+      "grad_norm": 0.5109714865684509,
+      "learning_rate": 6.645100322336118e-09,
+      "loss": 0.6359,
+      "step": 934
+    },
+    {
+      "epoch": 2.9526315789473685,
+      "grad_norm": 0.517336368560791,
+      "learning_rate": 5.7298787296750094e-09,
+      "loss": 0.6493,
+      "step": 935
+    },
+    {
+      "epoch": 2.9557894736842103,
+      "grad_norm": 0.5175721049308777,
+      "learning_rate": 4.88240150885555e-09,
+      "loss": 0.6275,
+      "step": 936
+    },
+    {
+      "epoch": 2.9589473684210525,
+      "grad_norm": 0.52158522605896,
+      "learning_rate": 4.1026801554139296e-09,
+      "loss": 0.613,
+      "step": 937
+    },
+    {
+      "epoch": 2.962105263157895,
+      "grad_norm": 0.5218417048454285,
+      "learning_rate": 3.3907252458176277e-09,
+      "loss": 0.6094,
+      "step": 938
+    },
+    {
+      "epoch": 2.9652631578947366,
+      "grad_norm": 0.5241231322288513,
+      "learning_rate": 2.7465464373205296e-09,
+      "loss": 0.6365,
+      "step": 939
+    },
+    {
+      "epoch": 2.968421052631579,
+      "grad_norm": 0.5028926730155945,
+      "learning_rate": 2.1701524678346964e-09,
+      "loss": 0.5889,
+      "step": 940
+    },
+    {
+      "epoch": 2.971578947368421,
+      "grad_norm": 0.5517033934593201,
+      "learning_rate": 1.6615511558082386e-09,
+      "loss": 0.559,
+      "step": 941
+    },
+    {
+      "epoch": 2.974736842105263,
+      "grad_norm": 0.49616527557373047,
+      "learning_rate": 1.220749400123733e-09,
+      "loss": 0.6422,
+      "step": 942
+    },
+    {
+      "epoch": 2.977894736842105,
+      "grad_norm": 0.5219214558601379,
+      "learning_rate": 8.477531799999661e-10,
+      "loss": 0.6043,
+      "step": 943
+    },
+    {
+      "epoch": 2.9810526315789474,
+      "grad_norm": 0.5003503561019897,
+      "learning_rate": 5.425675549136645e-10,
+      "loss": 0.6027,
+      "step": 944
+    },
+    {
+      "epoch": 2.984210526315789,
+      "grad_norm": 0.5330973863601685,
+      "learning_rate": 3.051966645312154e-10,
+      "loss": 0.5806,
+      "step": 945
+    },
+    {
+      "epoch": 2.9873684210526315,
+      "grad_norm": 0.5317094326019287,
+      "learning_rate": 1.3564372865038088e-10,
+      "loss": 0.607,
+      "step": 946
+    },
+    {
+      "epoch": 2.9905263157894737,
+      "grad_norm": 0.5060402750968933,
+      "learning_rate": 3.3911047158663445e-11,
+      "loss": 0.6126,
+      "step": 947
+    },
+    {
+      "epoch": 2.993684210526316,
+      "grad_norm": 0.5056828260421753,
+      "learning_rate": 0.0,
+      "loss": 0.6224,
+      "step": 948
+    },
+    {
+      "epoch": 2.993684210526316,
+      "step": 948,
+      "total_flos": 8.371731904995656e+17,
+      "train_loss": 0.7066250008998541,
+      "train_runtime": 24790.5767,
+      "train_samples_per_second": 3.677,
+      "train_steps_per_second": 0.038
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 948,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.371731904995656e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}