diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5060 @@
+{
+  "best_global_step": 700,
+  "best_metric": 2.0324764251708984,
+  "best_model_checkpoint": "./output_dir/fr-Llama-3.1-8B-lr4e-06-atten0.25-ffn0.25_20250430_122245/checkpoint-700",
+  "epoch": 0.41999160016799664,
+  "eval_steps": 50,
+  "global_step": 700,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005999880002399952,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0,
+      "loss": 1.9517,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011999760004799903,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8e-08,
+      "loss": 1.8084,
+      "step": 2
+    },
+    {
+      "epoch": 0.0017999640007199855,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6e-07,
+      "loss": 1.8214,
+      "step": 3
+    },
+    {
+      "epoch": 0.0023999520009599807,
+      "grad_norm": 1.7109375,
+      "learning_rate": 2.4e-07,
+      "loss": 1.8709,
+      "step": 4
+    },
+    {
+      "epoch": 0.002999940001199976,
+      "grad_norm": 2.125,
+      "learning_rate": 3.2e-07,
+      "loss": 1.8741,
+      "step": 5
+    },
+    {
+      "epoch": 0.003599928001439971,
+      "grad_norm": 2.015625,
+      "learning_rate": 4e-07,
+      "loss": 1.8875,
+      "step": 6
+    },
+    {
+      "epoch": 0.004199916001679967,
+      "grad_norm": 1.8359375,
+      "learning_rate": 4.8e-07,
+      "loss": 1.9111,
+      "step": 7
+    },
+    {
+      "epoch": 0.004799904001919961,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.6e-07,
+      "loss": 1.9349,
+      "step": 8
+    },
+    {
+      "epoch": 0.005399892002159957,
+      "grad_norm": 1.8671875,
+      "learning_rate": 6.4e-07,
+      "loss": 1.891,
+      "step": 9
+    },
+    {
+      "epoch": 0.005999880002399952,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.2e-07,
+      "loss": 1.9387,
+      "step": 10
+    },
+    {
+      "epoch": 0.006599868002639947,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8e-07,
+      "loss": 1.8719,
+      "step": 11
+    },
+    {
+      "epoch": 0.007199856002879942,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.799999999999999e-07,
+      "loss": 2.0476,
+      "step": 12
+    },
+    {
+      "epoch": 0.007799844003119938,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.6e-07,
+      "loss": 1.8709,
+      "step": 13
+    },
+    {
+      "epoch": 0.008399832003359933,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.04e-06,
+      "loss": 1.83,
+      "step": 14
+    },
+    {
+      "epoch": 0.008999820003599928,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.12e-06,
+      "loss": 1.8584,
+      "step": 15
+    },
+    {
+      "epoch": 0.009599808003839923,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.2e-06,
+      "loss": 1.9224,
+      "step": 16
+    },
+    {
+      "epoch": 0.01019979600407992,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.28e-06,
+      "loss": 1.8243,
+      "step": 17
+    },
+    {
+      "epoch": 0.010799784004319914,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.3600000000000001e-06,
+      "loss": 1.9473,
+      "step": 18
+    },
+    {
+      "epoch": 0.011399772004559909,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.44e-06,
+      "loss": 1.9006,
+      "step": 19
+    },
+    {
+      "epoch": 0.011999760004799903,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5199999999999998e-06,
+      "loss": 1.8361,
+      "step": 20
+    },
+    {
+      "epoch": 0.0125997480050399,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.6e-06,
+      "loss": 1.9316,
+      "step": 21
+    },
+    {
+      "epoch": 0.013199736005279895,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.6799999999999998e-06,
+      "loss": 1.8865,
+      "step": 22
+    },
+    {
+      "epoch": 0.01379972400551989,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.7599999999999999e-06,
+      "loss": 1.9309,
+      "step": 23
+    },
+    {
+      "epoch": 0.014399712005759884,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.84e-06,
+      "loss": 1.8815,
+      "step": 24
+    },
+    {
+      "epoch": 0.01499970000599988,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.92e-06,
+      "loss": 1.8123,
+      "step": 25
+    },
+    {
+      "epoch": 0.015599688006239875,
+      "grad_norm": 1.703125,
+      "learning_rate": 2e-06,
+      "loss": 1.8855,
+      "step": 26
+    },
+    {
+      "epoch": 0.016199676006479872,
+      "grad_norm": 2.0,
+      "learning_rate": 2.08e-06,
+      "loss": 1.8313,
+      "step": 27
+    },
+    {
+      "epoch": 0.016799664006719867,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.16e-06,
+      "loss": 1.8148,
+      "step": 28
+    },
+    {
+      "epoch": 0.01739965200695986,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.24e-06,
+      "loss": 1.9719,
+      "step": 29
+    },
+    {
+      "epoch": 0.017999640007199856,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.32e-06,
+      "loss": 1.8331,
+      "step": 30
+    },
+    {
+      "epoch": 0.01859962800743985,
+      "grad_norm": 1.890625,
+      "learning_rate": 2.4e-06,
+      "loss": 1.9469,
+      "step": 31
+    },
+    {
+      "epoch": 0.019199616007679846,
+      "grad_norm": 1.6953125,
+      "learning_rate": 2.48e-06,
+      "loss": 1.884,
+      "step": 32
+    },
+    {
+      "epoch": 0.01979960400791984,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.56e-06,
+      "loss": 1.9041,
+      "step": 33
+    },
+    {
+      "epoch": 0.02039959200815984,
+      "grad_norm": 1.9609375,
+      "learning_rate": 2.64e-06,
+      "loss": 2.0232,
+      "step": 34
+    },
+    {
+      "epoch": 0.020999580008399833,
+      "grad_norm": 2.125,
+      "learning_rate": 2.7200000000000002e-06,
+      "loss": 1.832,
+      "step": 35
+    },
+    {
+      "epoch": 0.021599568008639828,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.8e-06,
+      "loss": 1.7751,
+      "step": 36
+    },
+    {
+      "epoch": 0.022199556008879823,
+      "grad_norm": 1.875,
+      "learning_rate": 2.88e-06,
+      "loss": 1.8467,
+      "step": 37
+    },
+    {
+      "epoch": 0.022799544009119817,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.96e-06,
+      "loss": 1.7807,
+      "step": 38
+    },
+    {
+      "epoch": 0.023399532009359812,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.0399999999999997e-06,
+      "loss": 1.7781,
+      "step": 39
+    },
+    {
+      "epoch": 0.023999520009599807,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.1199999999999998e-06,
+      "loss": 1.8486,
+      "step": 40
+    },
+    {
+      "epoch": 0.0245995080098398,
+      "grad_norm": 1.9296875,
+      "learning_rate": 3.2e-06,
+      "loss": 1.8328,
+      "step": 41
+    },
+    {
+      "epoch": 0.0251994960100798,
+      "grad_norm": 1.7265625,
+      "learning_rate": 3.2799999999999995e-06,
+      "loss": 2.0436,
+      "step": 42
+    },
+    {
+      "epoch": 0.025799484010319795,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.3599999999999996e-06,
+      "loss": 1.9575,
+      "step": 43
+    },
+    {
+      "epoch": 0.02639947201055979,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.4399999999999997e-06,
+      "loss": 1.9247,
+      "step": 44
+    },
+    {
+      "epoch": 0.026999460010799784,
+      "grad_norm": 1.734375,
+      "learning_rate": 3.5199999999999998e-06,
+      "loss": 1.9135,
+      "step": 45
+    },
+    {
+      "epoch": 0.02759944801103978,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.6e-06,
+      "loss": 1.9564,
+      "step": 46
+    },
+    {
+      "epoch": 0.028199436011279774,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.68e-06,
+      "loss": 1.8047,
+      "step": 47
+    },
+    {
+      "epoch": 0.02879942401151977,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.7599999999999996e-06,
+      "loss": 1.9223,
+      "step": 48
+    },
+    {
+      "epoch": 0.029399412011759767,
+      "grad_norm": 1.7109375,
+      "learning_rate": 3.84e-06,
+      "loss": 1.9724,
+      "step": 49
+    },
+    {
+      "epoch": 0.02999940001199976,
+      "grad_norm": 1.7890625,
+      "learning_rate": 3.92e-06,
+      "loss": 1.8151,
+      "step": 50
+    },
+    {
+      "epoch": 0.02999940001199976,
+      "eval_loss": 2.060624361038208,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.0119,
+      "eval_samples_per_second": 153.818,
+      "eval_steps_per_second": 25.641,
+      "step": 50
+    },
+    {
+      "epoch": 0.030599388012239756,
+      "grad_norm": 1.46875,
+      "learning_rate": 4e-06,
+      "loss": 1.8375,
+      "step": 51
+    },
+    {
+      "epoch": 0.03119937601247975,
+      "grad_norm": 1.7109375,
+      "learning_rate": 3.997524752475248e-06,
+      "loss": 1.8369,
+      "step": 52
+    },
+    {
+      "epoch": 0.031799364012719745,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.9950495049504945e-06,
+      "loss": 1.7909,
+      "step": 53
+    },
+    {
+      "epoch": 0.032399352012959744,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.992574257425742e-06,
+      "loss": 1.819,
+      "step": 54
+    },
+    {
+      "epoch": 0.032999340013199735,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.99009900990099e-06,
+      "loss": 1.8014,
+      "step": 55
+    },
+    {
+      "epoch": 0.03359932801343973,
+      "grad_norm": 1.96875,
+      "learning_rate": 3.987623762376238e-06,
+      "loss": 1.7094,
+      "step": 56
+    },
+    {
+      "epoch": 0.034199316013679724,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.985148514851485e-06,
+      "loss": 1.8507,
+      "step": 57
+    },
+    {
+      "epoch": 0.03479930401391972,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.9826732673267325e-06,
+      "loss": 1.8693,
+      "step": 58
+    },
+    {
+      "epoch": 0.035399292014159714,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.98019801980198e-06,
+      "loss": 1.9363,
+      "step": 59
+    },
+    {
+      "epoch": 0.03599928001439971,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.977722772277228e-06,
+      "loss": 1.8308,
+      "step": 60
+    },
+    {
+      "epoch": 0.03659926801463971,
+      "grad_norm": 1.71875,
+      "learning_rate": 3.975247524752475e-06,
+      "loss": 1.8361,
+      "step": 61
+    },
+    {
+      "epoch": 0.0371992560148797,
+      "grad_norm": 1.375,
+      "learning_rate": 3.972772277227723e-06,
+      "loss": 1.839,
+      "step": 62
+    },
+    {
+      "epoch": 0.0377992440151197,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.9702970297029705e-06,
+      "loss": 1.8829,
+      "step": 63
+    },
+    {
+      "epoch": 0.03839923201535969,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.967821782178218e-06,
+      "loss": 1.88,
+      "step": 64
+    },
+    {
+      "epoch": 0.03899922001559969,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.965346534653465e-06,
+      "loss": 1.9815,
+      "step": 65
+    },
+    {
+      "epoch": 0.03959920801583968,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.962871287128713e-06,
+      "loss": 1.8474,
+      "step": 66
+    },
+    {
+      "epoch": 0.04019919601607968,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.96039603960396e-06,
+      "loss": 1.8953,
+      "step": 67
+    },
+    {
+      "epoch": 0.04079918401631968,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.957920792079208e-06,
+      "loss": 1.861,
+      "step": 68
+    },
+    {
+      "epoch": 0.04139917201655967,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.955445544554455e-06,
+      "loss": 1.9094,
+      "step": 69
+    },
+    {
+      "epoch": 0.041999160016799666,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.952970297029703e-06,
+      "loss": 2.0281,
+      "step": 70
+    },
+    {
+      "epoch": 0.04259914801703966,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.95049504950495e-06,
+      "loss": 1.896,
+      "step": 71
+    },
+    {
+      "epoch": 0.043199136017279656,
+      "grad_norm": 1.796875,
+      "learning_rate": 3.948019801980198e-06,
+      "loss": 1.8792,
+      "step": 72
+    },
+    {
+      "epoch": 0.04379912401751965,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.945544554455446e-06,
+      "loss": 1.8139,
+      "step": 73
+    },
+    {
+      "epoch": 0.044399112017759645,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.943069306930693e-06,
+      "loss": 1.813,
+      "step": 74
+    },
+    {
+      "epoch": 0.04499910001799964,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.94059405940594e-06,
+      "loss": 1.8587,
+      "step": 75
+    },
+    {
+      "epoch": 0.045599088018239635,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.938118811881188e-06,
+      "loss": 1.8793,
+      "step": 76
+    },
+    {
+      "epoch": 0.04619907601847963,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.935643564356436e-06,
+      "loss": 1.7305,
+      "step": 77
+    },
+    {
+      "epoch": 0.046799064018719624,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.933168316831683e-06,
+      "loss": 1.8243,
+      "step": 78
+    },
+    {
+      "epoch": 0.04739905201895962,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.9306930693069305e-06,
+      "loss": 1.9018,
+      "step": 79
+    },
+    {
+      "epoch": 0.047999040019199614,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.928217821782178e-06,
+      "loss": 1.8979,
+      "step": 80
+    },
+    {
+      "epoch": 0.04859902801943961,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.925742574257425e-06,
+      "loss": 1.9308,
+      "step": 81
+    },
+    {
+      "epoch": 0.0491990160196796,
+      "grad_norm": 1.6796875,
+      "learning_rate": 3.923267326732673e-06,
+      "loss": 1.8194,
+      "step": 82
+    },
+    {
+      "epoch": 0.0497990040199196,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.920792079207921e-06,
+      "loss": 1.8127,
+      "step": 83
+    },
+    {
+      "epoch": 0.0503989920201596,
+      "grad_norm": 1.625,
+      "learning_rate": 3.9183168316831685e-06,
+      "loss": 1.7168,
+      "step": 84
+    },
+    {
+      "epoch": 0.05099898002039959,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.915841584158415e-06,
+      "loss": 1.9451,
+      "step": 85
+    },
+    {
+      "epoch": 0.05159896802063959,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.913366336633663e-06,
+      "loss": 1.8582,
+      "step": 86
+    },
+    {
+      "epoch": 0.05219895602087958,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.910891089108911e-06,
+      "loss": 1.7901,
+      "step": 87
+    },
+    {
+      "epoch": 0.05279894402111958,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.908415841584159e-06,
+      "loss": 1.7489,
+      "step": 88
+    },
+    {
+      "epoch": 0.05339893202135957,
+      "grad_norm": 1.625,
+      "learning_rate": 3.905940594059406e-06,
+      "loss": 1.8365,
+      "step": 89
+    },
+    {
+      "epoch": 0.05399892002159957,
+      "grad_norm": 1.375,
+      "learning_rate": 3.903465346534653e-06,
+      "loss": 1.8664,
+      "step": 90
+    },
+    {
+      "epoch": 0.054598908021839566,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.9009900990099e-06,
+      "loss": 1.7951,
+      "step": 91
+    },
+    {
+      "epoch": 0.05519889602207956,
+      "grad_norm": 1.7421875,
+      "learning_rate": 3.898514851485148e-06,
+      "loss": 1.6784,
+      "step": 92
+    },
+    {
+      "epoch": 0.055798884022319556,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.896039603960396e-06,
+      "loss": 1.8358,
+      "step": 93
+    },
+    {
+      "epoch": 0.05639887202255955,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.893564356435644e-06,
+      "loss": 1.7865,
+      "step": 94
+    },
+    {
+      "epoch": 0.056998860022799545,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.8910891089108905e-06,
+      "loss": 1.8682,
+      "step": 95
+    },
+    {
+      "epoch": 0.05759884802303954,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.888613861386138e-06,
+      "loss": 1.8253,
+      "step": 96
+    },
+    {
+      "epoch": 0.058198836023279535,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.886138613861386e-06,
+      "loss": 1.7567,
+      "step": 97
+    },
+    {
+      "epoch": 0.05879882402351953,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.883663366336634e-06,
+      "loss": 2.0378,
+      "step": 98
+    },
+    {
+      "epoch": 0.059398812023759524,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.881188118811881e-06,
+      "loss": 1.9027,
+      "step": 99
+    },
+    {
+      "epoch": 0.05999880002399952,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.8787128712871285e-06,
+      "loss": 2.0289,
+      "step": 100
+    },
+    {
+      "epoch": 0.05999880002399952,
+      "eval_loss": 2.0494163036346436,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.139,
+      "eval_samples_per_second": 153.518,
+      "eval_steps_per_second": 25.591,
+      "step": 100
+    },
+    {
+      "epoch": 0.060598788024239514,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.876237623762376e-06,
+      "loss": 1.8943,
+      "step": 101
+    },
+    {
+      "epoch": 0.06119877602447951,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.873762376237624e-06,
+      "loss": 1.8859,
+      "step": 102
+    },
+    {
+      "epoch": 0.0617987640247195,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.871287128712871e-06,
+      "loss": 1.8459,
+      "step": 103
+    },
+    {
+      "epoch": 0.0623987520249595,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.868811881188119e-06,
+      "loss": 1.9159,
+      "step": 104
+    },
+    {
+      "epoch": 0.06299874002519949,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.866336633663366e-06,
+      "loss": 1.8295,
+      "step": 105
+    },
+    {
+      "epoch": 0.06359872802543949,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.8638613861386134e-06,
+      "loss": 1.9225,
+      "step": 106
+    },
+    {
+      "epoch": 0.06419871602567949,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.861386138613861e-06,
+      "loss": 1.8648,
+      "step": 107
+    },
+    {
+      "epoch": 0.06479870402591949,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.858910891089109e-06,
+      "loss": 1.8594,
+      "step": 108
+    },
+    {
+      "epoch": 0.06539869202615947,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.856435643564356e-06,
+      "loss": 1.979,
+      "step": 109
+    },
+    {
+      "epoch": 0.06599868002639947,
+      "grad_norm": 1.7734375,
+      "learning_rate": 3.853960396039604e-06,
+      "loss": 1.8235,
+      "step": 110
+    },
+    {
+      "epoch": 0.06659866802663947,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.851485148514851e-06,
+      "loss": 1.7781,
+      "step": 111
+    },
+    {
+      "epoch": 0.06719865602687947,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.849009900990099e-06,
+      "loss": 1.8508,
+      "step": 112
+    },
+    {
+      "epoch": 0.06779864402711945,
+      "grad_norm": 1.25,
+      "learning_rate": 3.846534653465346e-06,
+      "loss": 1.8538,
+      "step": 113
+    },
+    {
+      "epoch": 0.06839863202735945,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.844059405940594e-06,
+      "loss": 1.7364,
+      "step": 114
+    },
+    {
+      "epoch": 0.06899862002759945,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.841584158415842e-06,
+      "loss": 1.7373,
+      "step": 115
+    },
+    {
+      "epoch": 0.06959860802783945,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.839108910891089e-06,
+      "loss": 1.8901,
+      "step": 116
+    },
+    {
+      "epoch": 0.07019859602807944,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.836633663366336e-06,
+      "loss": 1.8541,
+      "step": 117
+    },
+    {
+      "epoch": 0.07079858402831943,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.834158415841584e-06,
+      "loss": 1.7285,
+      "step": 118
+    },
+    {
+      "epoch": 0.07139857202855943,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.831683168316831e-06,
+      "loss": 1.7322,
+      "step": 119
+    },
+    {
+      "epoch": 0.07199856002879942,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.829207920792079e-06,
+      "loss": 1.8256,
+      "step": 120
+    },
+    {
+      "epoch": 0.07259854802903942,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.8267326732673265e-06,
+      "loss": 1.7352,
+      "step": 121
+    },
+    {
+      "epoch": 0.07319853602927942,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.824257425742574e-06,
+      "loss": 1.8343,
+      "step": 122
+    },
+    {
+      "epoch": 0.0737985240295194,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.821782178217821e-06,
+      "loss": 1.9866,
+      "step": 123
+    },
+    {
+      "epoch": 0.0743985120297594,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.819306930693069e-06,
+      "loss": 1.8854,
+      "step": 124
+    },
+    {
+      "epoch": 0.0749985000299994,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.816831683168317e-06,
+      "loss": 1.8341,
+      "step": 125
+    },
+    {
+      "epoch": 0.0755984880302394,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.814356435643564e-06,
+      "loss": 1.829,
+      "step": 126
+    },
+    {
+      "epoch": 0.07619847603047938,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.8118811881188114e-06,
+      "loss": 1.8042,
+      "step": 127
+    },
+    {
+      "epoch": 0.07679846403071938,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.809405940594059e-06,
+      "loss": 1.7094,
+      "step": 128
+    },
+    {
+      "epoch": 0.07739845203095938,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.8069306930693065e-06,
+      "loss": 1.7705,
+      "step": 129
+    },
+    {
+      "epoch": 0.07799844003119938,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.8044554455445543e-06,
+      "loss": 1.8023,
+      "step": 130
+    },
+    {
+      "epoch": 0.07859842803143938,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.8019801980198017e-06,
+      "loss": 1.9408,
+      "step": 131
+    },
+    {
+      "epoch": 0.07919841603167936,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.7995049504950494e-06,
+      "loss": 1.8589,
+      "step": 132
+    },
+    {
+      "epoch": 0.07979840403191936,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.7970297029702968e-06,
+      "loss": 1.8806,
+      "step": 133
+    },
+    {
+      "epoch": 0.08039839203215936,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.7945544554455445e-06,
+      "loss": 1.8629,
+      "step": 134
+    },
+    {
+      "epoch": 0.08099838003239936,
+      "grad_norm": 1.734375,
+      "learning_rate": 3.792079207920792e-06,
+      "loss": 1.7811,
+      "step": 135
+    },
+    {
+      "epoch": 0.08159836803263935,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.7896039603960396e-06,
+      "loss": 1.8703,
+      "step": 136
+    },
+    {
+      "epoch": 0.08219835603287934,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.7871287128712866e-06,
+      "loss": 1.9697,
+      "step": 137
+    },
+    {
+      "epoch": 0.08279834403311934,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.7846534653465343e-06,
+      "loss": 1.8083,
+      "step": 138
+    },
+    {
+      "epoch": 0.08339833203335933,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.7821782178217817e-06,
+      "loss": 1.8348,
+      "step": 139
+    },
+    {
+      "epoch": 0.08399832003359933,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.7797029702970294e-06,
+      "loss": 1.7378,
+      "step": 140
+    },
+    {
+      "epoch": 0.08459830803383932,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.7772277227722768e-06,
+      "loss": 1.9362,
+      "step": 141
+    },
+    {
+      "epoch": 0.08519829603407932,
+      "grad_norm": 1.75,
+      "learning_rate": 3.7747524752475245e-06,
+      "loss": 1.9628,
+      "step": 142
+    },
+    {
+      "epoch": 0.08579828403431931,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.772277227722772e-06,
+      "loss": 1.8524,
+      "step": 143
+    },
+    {
+      "epoch": 0.08639827203455931,
+      "grad_norm": 1.6875,
+      "learning_rate": 3.7698019801980197e-06,
+      "loss": 1.7373,
+      "step": 144
+    },
+    {
+      "epoch": 0.08699826003479931,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.767326732673267e-06,
+      "loss": 1.8752,
+      "step": 145
+    },
+    {
+      "epoch": 0.0875982480350393,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.7648514851485148e-06,
+      "loss": 1.7939,
+      "step": 146
+    },
+    {
+      "epoch": 0.08819823603527929,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.762376237623762e-06,
+      "loss": 1.8857,
+      "step": 147
+    },
+    {
+      "epoch": 0.08879822403551929,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.75990099009901e-06,
+      "loss": 1.8343,
+      "step": 148
+    },
+    {
+      "epoch": 0.08939821203575929,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.7574257425742572e-06,
+      "loss": 1.8464,
+      "step": 149
+    },
+    {
+      "epoch": 0.08999820003599927,
+      "grad_norm": 1.375,
+      "learning_rate": 3.754950495049505e-06,
+      "loss": 1.6997,
+      "step": 150
+    },
+    {
+      "epoch": 0.08999820003599927,
+      "eval_loss": 2.043649673461914,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 67.4909,
+      "eval_samples_per_second": 148.168,
+      "eval_steps_per_second": 24.7,
+      "step": 150
+    },
+    {
+      "epoch": 0.09059818803623927,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.752475247524752e-06,
+      "loss": 1.8642,
+      "step": 151
+    },
+    {
+      "epoch": 0.09119817603647927,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.7499999999999997e-06,
+      "loss": 1.7887,
+      "step": 152
+    },
+    {
+      "epoch": 0.09179816403671927,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.7475247524752474e-06,
+      "loss": 1.8596,
+      "step": 153
+    },
+    {
+      "epoch": 0.09239815203695927,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.7450495049504948e-06,
+      "loss": 2.0243,
+      "step": 154
+    },
+    {
+      "epoch": 0.09299814003719925,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.7425742574257425e-06,
+      "loss": 1.7887,
+      "step": 155
+    },
+    {
+      "epoch": 0.09359812803743925,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.74009900990099e-06,
+      "loss": 1.828,
+      "step": 156
+    },
+    {
+      "epoch": 0.09419811603767925,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.7376237623762377e-06,
+      "loss": 1.7897,
+      "step": 157
+    },
+    {
+      "epoch": 0.09479810403791925,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.735148514851485e-06,
+      "loss": 1.761,
+      "step": 158
+    },
+    {
+      "epoch": 0.09539809203815924,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.7326732673267328e-06,
+      "loss": 1.812,
+      "step": 159
+    },
+    {
+      "epoch": 0.09599808003839923,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.73019801980198e-06,
+      "loss": 1.8758,
+      "step": 160
+    },
+    {
+      "epoch": 0.09659806803863923,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.727722772277228e-06,
+      "loss": 1.7899,
+      "step": 161
+    },
+    {
+      "epoch": 0.09719805603887922,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.7252475247524752e-06,
+      "loss": 1.7542,
+      "step": 162
+    },
+    {
+      "epoch": 0.09779804403911922,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.722772277227723e-06,
+      "loss": 1.765,
+      "step": 163
+    },
+    {
+      "epoch": 0.0983980320393592,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.72029702970297e-06,
+      "loss": 1.7773,
+      "step": 164
+    },
+    {
+      "epoch": 0.0989980200395992,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.7178217821782177e-06,
+      "loss": 1.9011,
+      "step": 165
+    },
+    {
+      "epoch": 0.0995980080398392,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.715346534653465e-06,
+      "loss": 1.8742,
+      "step": 166
+    },
+    {
+      "epoch": 0.1001979960400792,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.7128712871287128e-06,
+      "loss": 1.8805,
+      "step": 167
+    },
+    {
+      "epoch": 0.1007979840403192,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.71039603960396e-06,
+      "loss": 1.9764,
+      "step": 168
+    },
+    {
+      "epoch": 0.10139797204055918,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.707920792079208e-06,
+      "loss": 1.8004,
+      "step": 169
+    },
+    {
+      "epoch": 0.10199796004079918,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.7054455445544552e-06,
+      "loss": 1.8625,
+      "step": 170
+    },
+    {
+      "epoch": 0.10259794804103918,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.702970297029703e-06,
+      "loss": 1.7509,
+      "step": 171
+    },
+    {
+      "epoch": 0.10319793604127918,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.7004950495049503e-06,
+      "loss": 1.7476,
+      "step": 172
+    },
+    {
+      "epoch": 0.10379792404151916,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.698019801980198e-06,
+      "loss": 1.8179,
+      "step": 173
+    },
+    {
+      "epoch": 0.10439791204175916,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.6955445544554455e-06,
+      "loss": 1.9345,
+      "step": 174
+    },
+    {
+      "epoch": 0.10499790004199916,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.6930693069306932e-06,
+      "loss": 1.8446,
+      "step": 175
+    },
+    {
+      "epoch": 0.10559788804223916,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.6905940594059406e-06,
+      "loss": 1.9043,
+      "step": 176
+    },
+    {
+      "epoch": 0.10619787604247916,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.6881188118811883e-06,
+      "loss": 1.8542,
+      "step": 177
+    },
+    {
+      "epoch": 0.10679786404271914,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.6856435643564352e-06,
+      "loss": 1.9043,
+      "step": 178
+    },
+    {
+      "epoch": 0.10739785204295914,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.683168316831683e-06,
+      "loss": 1.8059,
+      "step": 179
+    },
+    {
+      "epoch": 0.10799784004319914,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.6806930693069304e-06,
+      "loss": 1.6908,
+      "step": 180
+    },
+    {
+      "epoch": 0.10859782804343913,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.678217821782178e-06,
+      "loss": 1.8294,
+      "step": 181
+    },
+    {
+      "epoch": 0.10919781604367913,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.6757425742574255e-06,
+      "loss": 2.0129,
+      "step": 182
+    },
+    {
+      "epoch": 0.10979780404391912,
+      "grad_norm": 1.5,
+      "learning_rate": 3.6732673267326732e-06,
+      "loss": 1.7686,
+      "step": 183
+    },
+    {
+      "epoch": 0.11039779204415912,
+      "grad_norm": 1.7421875,
+      "learning_rate": 3.6707920792079206e-06,
+      "loss": 1.9643,
+      "step": 184
+    },
+    {
+      "epoch": 0.11099778004439911,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.6683168316831683e-06,
+      "loss": 1.7562,
+      "step": 185
+    },
+    {
+      "epoch": 0.11159776804463911,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.6658415841584157e-06,
+      "loss": 1.8898,
+      "step": 186
+    },
+    {
+      "epoch": 0.1121977560448791,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.6633663366336635e-06,
+      "loss": 1.885,
+      "step": 187
+    },
+    {
+      "epoch": 0.1127977440451191,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.660891089108911e-06,
+      "loss": 1.9108,
+      "step": 188
+    },
+    {
+      "epoch": 0.11339773204535909,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.6584158415841586e-06,
+      "loss": 1.826,
+      "step": 189
+    },
+    {
+      "epoch": 0.11399772004559909,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.6559405940594055e-06,
+      "loss": 1.8905,
+      "step": 190
+    },
+    {
+      "epoch": 0.11459770804583909,
+      "grad_norm": 1.25,
+      "learning_rate": 3.6534653465346532e-06,
+      "loss": 1.8051,
+      "step": 191
+    },
+    {
+      "epoch": 0.11519769604607907,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.6509900990099006e-06,
+      "loss": 1.7055,
+      "step": 192
+    },
+    {
+      "epoch": 0.11579768404631907,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.6485148514851484e-06,
+      "loss": 1.8768,
+      "step": 193
+    },
+    {
+      "epoch": 0.11639767204655907,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.6460396039603957e-06,
+      "loss": 1.7644,
+      "step": 194
+    },
+    {
+      "epoch": 0.11699766004679907,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.6435643564356435e-06,
+      "loss": 1.9606,
+      "step": 195
+    },
+    {
+      "epoch": 0.11759764804703907,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.641089108910891e-06,
+      "loss": 1.7722,
+      "step": 196
+    },
+    {
+      "epoch": 0.11819763604727905,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.6386138613861386e-06,
+      "loss": 1.7471,
+      "step": 197
+    },
+    {
+      "epoch": 0.11879762404751905,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.636138613861386e-06,
+      "loss": 1.8394,
+      "step": 198
+    },
+    {
+      "epoch": 0.11939761204775905,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.6336633663366337e-06,
+      "loss": 1.8986,
+      "step": 199
+    },
+    {
+      "epoch": 0.11999760004799905,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.631188118811881e-06,
+      "loss": 1.6904,
+      "step": 200
+    },
+    {
+      "epoch": 0.11999760004799905,
+      "eval_loss": 2.040451765060425,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.1419,
+      "eval_samples_per_second": 153.511,
+      "eval_steps_per_second": 25.59,
+      "step": 200
+    },
+    {
+      "epoch": 0.12059758804823903,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.628712871287129e-06,
+      "loss": 1.8726,
+      "step": 201
+    },
+    {
+      "epoch": 0.12119757604847903,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.626237623762376e-06,
+      "loss": 1.8508,
+      "step": 202
+    },
+    {
+      "epoch": 0.12179756404871903,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.623762376237624e-06,
+      "loss": 1.9293,
+      "step": 203
+    },
+    {
+      "epoch": 0.12239755204895902,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.621287128712871e-06,
+      "loss": 1.9013,
+      "step": 204
+    },
+    {
+      "epoch": 0.12299754004919902,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.6188118811881186e-06,
+      "loss": 1.9189,
+      "step": 205
+    },
+    {
+      "epoch": 0.123597528049439,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.616336633663366e-06,
+      "loss": 1.8183,
+      "step": 206
+    },
+    {
+      "epoch": 0.124197516049679,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.6138613861386137e-06,
+      "loss": 1.8873,
+      "step": 207
+    },
+    {
+      "epoch": 0.124797504049919,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.611386138613861e-06,
+      "loss": 1.8153,
+      "step": 208
+    },
+    {
+      "epoch": 0.125397492050159,
+      "grad_norm": 1.375,
+      "learning_rate": 3.608910891089109e-06,
+      "loss": 1.7048,
+      "step": 209
+    },
+    {
+      "epoch": 0.12599748005039899,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.606435643564356e-06,
+      "loss": 1.6584,
+      "step": 210
+    },
+    {
+      "epoch": 0.12659746805063898,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.603960396039604e-06,
+      "loss": 1.7825,
+      "step": 211
+    },
+    {
+      "epoch": 0.12719745605087898,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.6014851485148513e-06,
+      "loss": 1.8709,
+      "step": 212
+    },
+    {
+      "epoch": 0.12779744405111898,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.599009900990099e-06,
+      "loss": 1.858,
+      "step": 213
+    },
+    {
+      "epoch": 0.12839743205135898,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.5965346534653464e-06,
+      "loss": 1.9115,
+      "step": 214
+    },
+    {
+      "epoch": 0.12899742005159898,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.594059405940594e-06,
+      "loss": 1.7853,
+      "step": 215
+    },
+    {
+      "epoch": 0.12959740805183897,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.5915841584158415e-06,
+      "loss": 1.8218,
+      "step": 216
+    },
+    {
+      "epoch": 0.13019739605207895,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.589108910891089e-06,
+      "loss": 1.7609,
+      "step": 217
+    },
+    {
+      "epoch": 0.13079738405231894,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.586633663366336e-06,
+      "loss": 2.1418,
+      "step": 218
+    },
+    {
+      "epoch": 0.13139737205255894,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.584158415841584e-06,
+      "loss": 1.7065,
+      "step": 219
+    },
+    {
+      "epoch": 0.13199736005279894,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.5816831683168313e-06,
+      "loss": 1.8543,
+      "step": 220
+    },
+    {
+      "epoch": 0.13259734805303894,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.579207920792079e-06,
+      "loss": 1.8026,
+      "step": 221
+    },
+    {
+      "epoch": 0.13319733605327894,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.5767326732673264e-06,
+      "loss": 1.7373,
+      "step": 222
+    },
+    {
+      "epoch": 0.13379732405351893,
+      "grad_norm": 1.65625,
+      "learning_rate": 3.574257425742574e-06,
+      "loss": 1.7388,
+      "step": 223
+    },
+    {
+      "epoch": 0.13439731205375893,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.5717821782178215e-06,
+      "loss": 1.8013,
+      "step": 224
+    },
+    {
+      "epoch": 0.13499730005399893,
+      "grad_norm": 1.375,
+      "learning_rate": 3.5693069306930693e-06,
+      "loss": 1.9171,
+      "step": 225
+    },
+    {
+      "epoch": 0.1355972880542389,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.5668316831683166e-06,
+      "loss": 1.9055,
+      "step": 226
+    },
+    {
+      "epoch": 0.1361972760544789,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.5643564356435644e-06,
+      "loss": 1.9024,
+      "step": 227
+    },
+    {
+      "epoch": 0.1367972640547189,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.5618811881188117e-06,
+      "loss": 1.8366,
+      "step": 228
+    },
+    {
+      "epoch": 0.1373972520549589,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.5594059405940595e-06,
+      "loss": 1.8003,
+      "step": 229
+    },
+    {
+      "epoch": 0.1379972400551989,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.5569306930693064e-06,
+      "loss": 1.9299,
+      "step": 230
+    },
+    {
+      "epoch": 0.1385972280554389,
+      "grad_norm": 1.25,
+      "learning_rate": 3.554455445544554e-06,
+      "loss": 1.8557,
+      "step": 231
+    },
+    {
+      "epoch": 0.1391972160556789,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.5519801980198015e-06,
+      "loss": 1.8823,
+      "step": 232
+    },
+    {
+      "epoch": 0.1397972040559189,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.5495049504950493e-06,
+      "loss": 1.891,
+      "step": 233
+    },
+    {
+      "epoch": 0.1403971920561589,
+      "grad_norm": 1.734375,
+      "learning_rate": 3.5470297029702966e-06,
+      "loss": 1.6939,
+      "step": 234
+    },
+    {
+      "epoch": 0.14099718005639889,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.5445544554455444e-06,
+      "loss": 1.8792,
+      "step": 235
+    },
+    {
+      "epoch": 0.14159716805663886,
+      "grad_norm": 1.25,
+      "learning_rate": 3.5420792079207917e-06,
+      "loss": 1.856,
+      "step": 236
+    },
+    {
+      "epoch": 0.14219715605687885,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.5396039603960395e-06,
+      "loss": 1.8268,
+      "step": 237
+    },
+    {
+      "epoch": 0.14279714405711885,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.537128712871287e-06,
+      "loss": 1.7116,
+      "step": 238
+    },
+    {
+      "epoch": 0.14339713205735885,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.5346534653465346e-06,
+      "loss": 1.922,
+      "step": 239
+    },
+    {
+      "epoch": 0.14399712005759885,
+      "grad_norm": 1.5,
+      "learning_rate": 3.532178217821782e-06,
+      "loss": 1.792,
+      "step": 240
+    },
+    {
+      "epoch": 0.14459710805783885,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.5297029702970297e-06,
+      "loss": 1.6891,
+      "step": 241
+    },
+    {
+      "epoch": 0.14519709605807885,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.527227722772277e-06,
+      "loss": 1.8681,
+      "step": 242
+    },
+    {
+      "epoch": 0.14579708405831884,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.524752475247525e-06,
+      "loss": 1.6985,
+      "step": 243
+    },
+    {
+      "epoch": 0.14639707205855884,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.5222772277227717e-06,
+      "loss": 1.7168,
+      "step": 244
+    },
+    {
+      "epoch": 0.1469970600587988,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.5198019801980195e-06,
+      "loss": 1.7725,
+      "step": 245
+    },
+    {
+      "epoch": 0.1475970480590388,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.517326732673267e-06,
+      "loss": 1.8263,
+      "step": 246
+    },
+    {
+      "epoch": 0.1481970360592788,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.5148514851485146e-06,
+      "loss": 1.791,
+      "step": 247
+    },
+    {
+      "epoch": 0.1487970240595188,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.512376237623762e-06,
+      "loss": 1.867,
+      "step": 248
+    },
+    {
+      "epoch": 0.1493970120597588,
+      "grad_norm": 1.6953125,
+      "learning_rate": 3.5099009900990097e-06,
+      "loss": 1.92,
+      "step": 249
+    },
+    {
+      "epoch": 0.1499970000599988,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.507425742574257e-06,
+      "loss": 1.8612,
+      "step": 250
+    },
+    {
+      "epoch": 0.1499970000599988,
+      "eval_loss": 2.0383219718933105,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.0682,
+      "eval_samples_per_second": 153.685,
+      "eval_steps_per_second": 25.619,
+      "step": 250
+    },
+    {
+      "epoch": 0.1505969880602388,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.504950495049505e-06,
+      "loss": 1.8017,
+      "step": 251
+    },
+    {
+      "epoch": 0.1511969760604788,
+      "grad_norm": 1.375,
+      "learning_rate": 3.502475247524752e-06,
+      "loss": 1.944,
+      "step": 252
+    },
+    {
+      "epoch": 0.1517969640607188,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.5e-06,
+      "loss": 1.7536,
+      "step": 253
+    },
+    {
+      "epoch": 0.15239695206095877,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.4975247524752477e-06,
+      "loss": 1.7696,
+      "step": 254
+    },
+    {
+      "epoch": 0.15299694006119877,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.495049504950495e-06,
+      "loss": 1.8545,
+      "step": 255
+    },
+    {
+      "epoch": 0.15359692806143876,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.492574257425743e-06,
+      "loss": 1.9009,
+      "step": 256
+    },
+    {
+      "epoch": 0.15419691606167876,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.4900990099009897e-06,
+      "loss": 1.915,
+      "step": 257
+    },
+    {
+      "epoch": 0.15479690406191876,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.4876237623762375e-06,
+      "loss": 1.7616,
+      "step": 258
+    },
+    {
+      "epoch": 0.15539689206215876,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.485148514851485e-06,
+      "loss": 1.843,
+      "step": 259
+    },
+    {
+      "epoch": 0.15599688006239876,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.4826732673267326e-06,
+      "loss": 1.8939,
+      "step": 260
+    },
+    {
+      "epoch": 0.15659686806263876,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.48019801980198e-06,
+      "loss": 1.8774,
+      "step": 261
+    },
+    {
+      "epoch": 0.15719685606287875,
+      "grad_norm": 1.25,
+      "learning_rate": 3.4777227722772277e-06,
+      "loss": 1.7766,
+      "step": 262
+    },
+    {
+      "epoch": 0.15779684406311872,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.475247524752475e-06,
+      "loss": 1.806,
+      "step": 263
+    },
+    {
+      "epoch": 0.15839683206335872,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.472772277227723e-06,
+      "loss": 1.8267,
+      "step": 264
+    },
+    {
+      "epoch": 0.15899682006359872,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.47029702970297e-06,
+      "loss": 1.8525,
+      "step": 265
+    },
+    {
+      "epoch": 0.15959680806383872,
+      "grad_norm": 1.6796875,
+      "learning_rate": 3.467821782178218e-06,
+      "loss": 1.7749,
+      "step": 266
+    },
+    {
+      "epoch": 0.16019679606407872,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.4653465346534653e-06,
+      "loss": 1.905,
+      "step": 267
+    },
+    {
+      "epoch": 0.16079678406431872,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.462871287128713e-06,
+      "loss": 1.7351,
+      "step": 268
+    },
+    {
+      "epoch": 0.1613967720645587,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.4603960396039604e-06,
+      "loss": 1.8969,
+      "step": 269
+    },
+    {
+      "epoch": 0.1619967600647987,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.4579207920792077e-06,
+      "loss": 1.7878,
+      "step": 270
+    },
+    {
+      "epoch": 0.1625967480650387,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.455445544554455e-06,
+      "loss": 1.84,
+      "step": 271
+    },
+    {
+      "epoch": 0.1631967360652787,
+      "grad_norm": 1.375,
+      "learning_rate": 3.452970297029703e-06,
+      "loss": 1.9861,
+      "step": 272
+    },
+    {
+      "epoch": 0.16379672406551868,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.45049504950495e-06,
+      "loss": 1.8173,
+      "step": 273
+    },
+    {
+      "epoch": 0.16439671206575868,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.448019801980198e-06,
+      "loss": 1.8155,
+      "step": 274
+    },
+    {
+      "epoch": 0.16499670006599867,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.4455445544554453e-06,
+      "loss": 1.7419,
+      "step": 275
+    },
+    {
+      "epoch": 0.16559668806623867,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.443069306930693e-06,
+      "loss": 1.7549,
+      "step": 276
+    },
+    {
+      "epoch": 0.16619667606647867,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.4405940594059404e-06,
+      "loss": 1.8526,
+      "step": 277
+    },
+    {
+      "epoch": 0.16679666406671867,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.438118811881188e-06,
+      "loss": 1.8459,
+      "step": 278
+    },
+    {
+      "epoch": 0.16739665206695867,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.4356435643564355e-06,
+      "loss": 1.8812,
+      "step": 279
+    },
+    {
+      "epoch": 0.16799664006719867,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.4331683168316833e-06,
+      "loss": 1.8331,
+      "step": 280
+    },
+    {
+      "epoch": 0.16859662806743866,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.4306930693069306e-06,
+      "loss": 1.965,
+      "step": 281
+    },
+    {
+      "epoch": 0.16919661606767863,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.4282178217821784e-06,
+      "loss": 1.9069,
+      "step": 282
+    },
+    {
+      "epoch": 0.16979660406791863,
+      "grad_norm": 1.890625,
+      "learning_rate": 3.4257425742574253e-06,
+      "loss": 1.7781,
+      "step": 283
+    },
+    {
+      "epoch": 0.17039659206815863,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.423267326732673e-06,
+      "loss": 1.9266,
+      "step": 284
+    },
+    {
+      "epoch": 0.17099658006839863,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.4207920792079204e-06,
+      "loss": 1.8376,
+      "step": 285
+    },
+    {
+      "epoch": 0.17159656806863863,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.418316831683168e-06,
+      "loss": 1.7678,
+      "step": 286
+    },
+    {
+      "epoch": 0.17219655606887863,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.4158415841584155e-06,
+      "loss": 1.8636,
+      "step": 287
+    },
+    {
+      "epoch": 0.17279654406911862,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.4133663366336633e-06,
+      "loss": 1.8833,
+      "step": 288
+    },
+    {
+      "epoch": 0.17339653206935862,
+      "grad_norm": 1.375,
+      "learning_rate": 3.4108910891089106e-06,
+      "loss": 1.7882,
+      "step": 289
+    },
+    {
+      "epoch": 0.17399652006959862,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.4084158415841584e-06,
+      "loss": 1.9467,
+      "step": 290
+    },
+    {
+      "epoch": 0.1745965080698386,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.4059405940594058e-06,
+      "loss": 1.7262,
+      "step": 291
+    },
+    {
+      "epoch": 0.1751964960700786,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.4034653465346535e-06,
+      "loss": 1.8879,
+      "step": 292
+    },
+    {
+      "epoch": 0.1757964840703186,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.400990099009901e-06,
+      "loss": 1.8687,
+      "step": 293
+    },
+    {
+      "epoch": 0.17639647207055859,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.3985148514851486e-06,
+      "loss": 1.7491,
+      "step": 294
+    },
+    {
+      "epoch": 0.17699646007079858,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.396039603960396e-06,
+      "loss": 1.939,
+      "step": 295
+    },
+    {
+      "epoch": 0.17759644807103858,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.3935643564356437e-06,
+      "loss": 1.8243,
+      "step": 296
+    },
+    {
+      "epoch": 0.17819643607127858,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.3910891089108907e-06,
+      "loss": 1.8302,
+      "step": 297
+    },
+    {
+      "epoch": 0.17879642407151858,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.3886138613861384e-06,
+      "loss": 1.8882,
+      "step": 298
+    },
+    {
+      "epoch": 0.17939641207175858,
+      "grad_norm": 1.375,
+      "learning_rate": 3.3861386138613858e-06,
+      "loss": 1.8015,
+      "step": 299
+    },
+    {
+      "epoch": 0.17999640007199855,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.3836633663366335e-06,
+      "loss": 1.9455,
+      "step": 300
+    },
+    {
+      "epoch": 0.17999640007199855,
+      "eval_loss": 2.0367591381073,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 67.518,
+      "eval_samples_per_second": 148.109,
+      "eval_steps_per_second": 24.69,
+      "step": 300
+    },
+    {
+      "epoch": 0.18059638807223855,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.381188118811881e-06,
+      "loss": 1.7753,
+      "step": 301
+    },
+    {
+      "epoch": 0.18119637607247854,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.3787128712871286e-06,
+      "loss": 1.9745,
+      "step": 302
+    },
+    {
+      "epoch": 0.18179636407271854,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.376237623762376e-06,
+      "loss": 1.7602,
+      "step": 303
+    },
+    {
+      "epoch": 0.18239635207295854,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.3737623762376238e-06,
+      "loss": 1.9368,
+      "step": 304
+    },
+    {
+      "epoch": 0.18299634007319854,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.371287128712871e-06,
+      "loss": 1.9062,
+      "step": 305
+    },
+    {
+      "epoch": 0.18359632807343854,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.368811881188119e-06,
+      "loss": 1.7124,
+      "step": 306
+    },
+    {
+      "epoch": 0.18419631607367853,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.366336633663366e-06,
+      "loss": 1.7911,
+      "step": 307
+    },
+    {
+      "epoch": 0.18479630407391853,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.363861386138614e-06,
+      "loss": 1.8183,
+      "step": 308
+    },
+    {
+      "epoch": 0.1853962920741585,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.3613861386138613e-06,
+      "loss": 1.8593,
+      "step": 309
+    },
+    {
+      "epoch": 0.1859962800743985,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.3589108910891087e-06,
+      "loss": 1.9335,
+      "step": 310
+    },
+    {
+      "epoch": 0.1865962680746385,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.356435643564356e-06,
+      "loss": 1.8858,
+      "step": 311
+    },
+    {
+      "epoch": 0.1871962560748785,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.3539603960396038e-06,
+      "loss": 1.9154,
+      "step": 312
+    },
+    {
+      "epoch": 0.1877962440751185,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.351485148514851e-06,
+      "loss": 1.7767,
+      "step": 313
+    },
+    {
+      "epoch": 0.1883962320753585,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.349009900990099e-06,
+      "loss": 1.8943,
+      "step": 314
+    },
+    {
+      "epoch": 0.1889962200755985,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.3465346534653462e-06,
+      "loss": 1.7825,
+      "step": 315
+    },
+    {
+      "epoch": 0.1895962080758385,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.344059405940594e-06,
+      "loss": 1.891,
+      "step": 316
+    },
+    {
+      "epoch": 0.1901961960760785,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.3415841584158413e-06,
+      "loss": 1.8893,
+      "step": 317
+    },
+    {
+      "epoch": 0.1907961840763185,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.339108910891089e-06,
+      "loss": 1.7661,
+      "step": 318
+    },
+    {
+      "epoch": 0.19139617207655846,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.3366336633663364e-06,
+      "loss": 1.7951,
+      "step": 319
+    },
+    {
+      "epoch": 0.19199616007679846,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.334158415841584e-06,
+      "loss": 1.825,
+      "step": 320
+    },
+    {
+      "epoch": 0.19259614807703845,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.3316831683168316e-06,
+      "loss": 1.7065,
+      "step": 321
+    },
+    {
+      "epoch": 0.19319613607727845,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.3292079207920793e-06,
+      "loss": 1.8174,
+      "step": 322
+    },
+    {
+      "epoch": 0.19379612407751845,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.3267326732673262e-06,
+      "loss": 1.8812,
+      "step": 323
+    },
+    {
+      "epoch": 0.19439611207775845,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.324257425742574e-06,
+      "loss": 1.8562,
+      "step": 324
+    },
+    {
+      "epoch": 0.19499610007799845,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.3217821782178213e-06,
+      "loss": 1.7252,
+      "step": 325
+    },
+    {
+      "epoch": 0.19559608807823844,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.319306930693069e-06,
+      "loss": 1.7466,
+      "step": 326
+    },
+    {
+      "epoch": 0.19619607607847844,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.3168316831683165e-06,
+      "loss": 1.7714,
+      "step": 327
+    },
+    {
+      "epoch": 0.1967960640787184,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.3143564356435642e-06,
+      "loss": 1.8033,
+      "step": 328
+    },
+    {
+      "epoch": 0.1973960520789584,
+      "grad_norm": 1.71875,
+      "learning_rate": 3.3118811881188116e-06,
+      "loss": 1.812,
+      "step": 329
+    },
+    {
+      "epoch": 0.1979960400791984,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.3094059405940593e-06,
+      "loss": 1.8678,
+      "step": 330
+    },
+    {
+      "epoch": 0.1985960280794384,
+      "grad_norm": 1.25,
+      "learning_rate": 3.3069306930693067e-06,
+      "loss": 1.9161,
+      "step": 331
+    },
+    {
+      "epoch": 0.1991960160796784,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.3044554455445544e-06,
+      "loss": 1.7642,
+      "step": 332
+    },
+    {
+      "epoch": 0.1997960040799184,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.3019801980198018e-06,
+      "loss": 1.8247,
+      "step": 333
+    },
+    {
+      "epoch": 0.2003959920801584,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.2995049504950496e-06,
+      "loss": 1.9492,
+      "step": 334
+    },
+    {
+      "epoch": 0.2009959800803984,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.297029702970297e-06,
+      "loss": 1.7759,
+      "step": 335
+    },
+    {
+      "epoch": 0.2015959680806384,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.2945544554455442e-06,
+      "loss": 1.8604,
+      "step": 336
+    },
+    {
+      "epoch": 0.20219595608087837,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.2920792079207916e-06,
+      "loss": 1.8555,
+      "step": 337
+    },
+    {
+      "epoch": 0.20279594408111837,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.2896039603960393e-06,
+      "loss": 1.7694,
+      "step": 338
+    },
+    {
+      "epoch": 0.20339593208135837,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.2871287128712867e-06,
+      "loss": 1.7972,
+      "step": 339
+    },
+    {
+      "epoch": 0.20399592008159836,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.2846534653465345e-06,
+      "loss": 1.902,
+      "step": 340
+    },
+    {
+      "epoch": 0.20459590808183836,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.282178217821782e-06,
+      "loss": 1.9065,
+      "step": 341
+    },
+    {
+      "epoch": 0.20519589608207836,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.2797029702970296e-06,
+      "loss": 1.7904,
+      "step": 342
+    },
+    {
+      "epoch": 0.20579588408231836,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.277227722772277e-06,
+      "loss": 1.7835,
+      "step": 343
+    },
+    {
+      "epoch": 0.20639587208255836,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.2747524752475247e-06,
+      "loss": 1.8314,
+      "step": 344
+    },
+    {
+      "epoch": 0.20699586008279836,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.272277227722772e-06,
+      "loss": 1.774,
+      "step": 345
+    },
+    {
+      "epoch": 0.20759584808303833,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.2698019801980198e-06,
+      "loss": 1.8417,
+      "step": 346
+    },
+    {
+      "epoch": 0.20819583608327832,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.267326732673267e-06,
+      "loss": 1.8123,
+      "step": 347
+    },
+    {
+      "epoch": 0.20879582408351832,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.264851485148515e-06,
+      "loss": 1.8038,
+      "step": 348
+    },
+    {
+      "epoch": 0.20939581208375832,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.262376237623762e-06,
+      "loss": 1.9785,
+      "step": 349
+    },
+    {
+      "epoch": 0.20999580008399832,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.2599009900990096e-06,
+      "loss": 1.756,
+      "step": 350
+    },
+    {
+      "epoch": 0.20999580008399832,
+      "eval_loss": 2.0354440212249756,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.0421,
+      "eval_samples_per_second": 153.747,
+      "eval_steps_per_second": 25.63,
+      "step": 350
+    },
+    {
+      "epoch": 0.21059578808423832,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.257425742574257e-06,
+      "loss": 1.7987,
+      "step": 351
+    },
+    {
+      "epoch": 0.21119577608447831,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.2549504950495047e-06,
+      "loss": 1.868,
+      "step": 352
+    },
+    {
+      "epoch": 0.2117957640847183,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.252475247524752e-06,
+      "loss": 1.945,
+      "step": 353
+    },
+    {
+      "epoch": 0.2123957520849583,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.25e-06,
+      "loss": 1.869,
+      "step": 354
+    },
+    {
+      "epoch": 0.2129957400851983,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.2475247524752476e-06,
+      "loss": 1.6786,
+      "step": 355
+    },
+    {
+      "epoch": 0.21359572808543828,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.245049504950495e-06,
+      "loss": 1.885,
+      "step": 356
+    },
+    {
+      "epoch": 0.21419571608567828,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.2425742574257427e-06,
+      "loss": 1.8202,
+      "step": 357
+    },
+    {
+      "epoch": 0.21479570408591828,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.24009900990099e-06,
+      "loss": 1.8872,
+      "step": 358
+    },
+    {
+      "epoch": 0.21539569208615827,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.2376237623762378e-06,
+      "loss": 1.8847,
+      "step": 359
+    },
+    {
+      "epoch": 0.21599568008639827,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.235148514851485e-06,
+      "loss": 1.769,
+      "step": 360
+    },
+    {
+      "epoch": 0.21659566808663827,
+      "grad_norm": 1.25,
+      "learning_rate": 3.232673267326733e-06,
+      "loss": 1.8455,
+      "step": 361
+    },
+    {
+      "epoch": 0.21719565608687827,
+      "grad_norm": 1.375,
+      "learning_rate": 3.2301980198019802e-06,
+      "loss": 1.8822,
+      "step": 362
+    },
+    {
+      "epoch": 0.21779564408711827,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.2277227722772276e-06,
+      "loss": 1.863,
+      "step": 363
+    },
+    {
+      "epoch": 0.21839563208735827,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.225247524752475e-06,
+      "loss": 1.8595,
+      "step": 364
+    },
+    {
+      "epoch": 0.21899562008759824,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.2227722772277227e-06,
+      "loss": 1.8043,
+      "step": 365
+    },
+    {
+      "epoch": 0.21959560808783823,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.22029702970297e-06,
+      "loss": 1.8976,
+      "step": 366
+    },
+    {
+      "epoch": 0.22019559608807823,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.217821782178218e-06,
+      "loss": 1.8178,
+      "step": 367
+    },
+    {
+      "epoch": 0.22079558408831823,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.215346534653465e-06,
+      "loss": 1.7554,
+      "step": 368
+    },
+    {
+      "epoch": 0.22139557208855823,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.212871287128713e-06,
+      "loss": 1.8248,
+      "step": 369
+    },
+    {
+      "epoch": 0.22199556008879823,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.2103960396039603e-06,
+      "loss": 1.93,
+      "step": 370
+    },
+    {
+      "epoch": 0.22259554808903823,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.207920792079208e-06,
+      "loss": 1.9148,
+      "step": 371
+    },
+    {
+      "epoch": 0.22319553608927822,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.2054455445544554e-06,
+      "loss": 1.8438,
+      "step": 372
+    },
+    {
+      "epoch": 0.22379552408951822,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.202970297029703e-06,
+      "loss": 1.7993,
+      "step": 373
+    },
+    {
+      "epoch": 0.2243955120897582,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.2004950495049505e-06,
+      "loss": 1.8202,
+      "step": 374
+    },
+    {
+      "epoch": 0.2249955000899982,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.1980198019801982e-06,
+      "loss": 1.8704,
+      "step": 375
+    },
+    {
+      "epoch": 0.2255954880902382,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.195544554455445e-06,
+      "loss": 1.8558,
+      "step": 376
+    },
+    {
+      "epoch": 0.2261954760904782,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.193069306930693e-06,
+      "loss": 2.0454,
+      "step": 377
+    },
+    {
+      "epoch": 0.22679546409071819,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.1905940594059403e-06,
+      "loss": 1.8577,
+      "step": 378
+    },
+    {
+      "epoch": 0.22739545209095818,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.188118811881188e-06,
+      "loss": 1.8602,
+      "step": 379
+    },
+    {
+      "epoch": 0.22799544009119818,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.1856435643564354e-06,
+      "loss": 1.7231,
+      "step": 380
+    },
+    {
+      "epoch": 0.22859542809143818,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.183168316831683e-06,
+      "loss": 1.712,
+      "step": 381
+    },
+    {
+      "epoch": 0.22919541609167818,
+      "grad_norm": 1.25,
+      "learning_rate": 3.1806930693069305e-06,
+      "loss": 1.8823,
+      "step": 382
+    },
+    {
+      "epoch": 0.22979540409191815,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.1782178217821783e-06,
+      "loss": 1.845,
+      "step": 383
+    },
+    {
+      "epoch": 0.23039539209215815,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.1757425742574256e-06,
+      "loss": 1.8086,
+      "step": 384
+    },
+    {
+      "epoch": 0.23099538009239814,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.1732673267326734e-06,
+      "loss": 1.8122,
+      "step": 385
+    },
+    {
+      "epoch": 0.23159536809263814,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.1707920792079207e-06,
+      "loss": 1.8456,
+      "step": 386
+    },
+    {
+      "epoch": 0.23219535609287814,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.1683168316831685e-06,
+      "loss": 1.8158,
+      "step": 387
+    },
+    {
+      "epoch": 0.23279534409311814,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.165841584158416e-06,
+      "loss": 1.8107,
+      "step": 388
+    },
+    {
+      "epoch": 0.23339533209335814,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.163366336633663e-06,
+      "loss": 1.825,
+      "step": 389
+    },
+    {
+      "epoch": 0.23399532009359814,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.1608910891089105e-06,
+      "loss": 1.8038,
+      "step": 390
+    },
+    {
+      "epoch": 0.23459530809383813,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.1584158415841583e-06,
+      "loss": 1.8152,
+      "step": 391
+    },
+    {
+      "epoch": 0.23519529609407813,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.1559405940594056e-06,
+      "loss": 1.926,
+      "step": 392
+    },
+    {
+      "epoch": 0.2357952840943181,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.1534653465346534e-06,
+      "loss": 1.8543,
+      "step": 393
+    },
+    {
+      "epoch": 0.2363952720945581,
+      "grad_norm": 1.25,
+      "learning_rate": 3.1509900990099007e-06,
+      "loss": 1.8222,
+      "step": 394
+    },
+    {
+      "epoch": 0.2369952600947981,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.1485148514851485e-06,
+      "loss": 1.9147,
+      "step": 395
+    },
+    {
+      "epoch": 0.2375952480950381,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.146039603960396e-06,
+      "loss": 1.814,
+      "step": 396
+    },
+    {
+      "epoch": 0.2381952360952781,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.1435643564356436e-06,
+      "loss": 1.8623,
+      "step": 397
+    },
+    {
+      "epoch": 0.2387952240955181,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.141089108910891e-06,
+      "loss": 1.8996,
+      "step": 398
+    },
+    {
+      "epoch": 0.2393952120957581,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.1386138613861387e-06,
+      "loss": 1.8078,
+      "step": 399
+    },
+    {
+      "epoch": 0.2399952000959981,
+      "grad_norm": 1.125,
+      "learning_rate": 3.136138613861386e-06,
+      "loss": 1.8248,
+      "step": 400
+    },
+    {
+      "epoch": 0.2399952000959981,
+      "eval_loss": 2.0347118377685547,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.1996,
+      "eval_samples_per_second": 153.375,
+      "eval_steps_per_second": 25.568,
+      "step": 400
+    },
+    {
+      "epoch": 0.2405951880962381,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.133663366336634e-06,
+      "loss": 1.9232,
+      "step": 401
+    },
+    {
+      "epoch": 0.24119517609647806,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.1311881188118807e-06,
+      "loss": 1.8487,
+      "step": 402
+    },
+    {
+      "epoch": 0.24179516409671806,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.1287128712871285e-06,
+      "loss": 1.8713,
+      "step": 403
+    },
+    {
+      "epoch": 0.24239515209695806,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.126237623762376e-06,
+      "loss": 1.9976,
+      "step": 404
+    },
+    {
+      "epoch": 0.24299514009719805,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.1237623762376236e-06,
+      "loss": 1.769,
+      "step": 405
+    },
+    {
+      "epoch": 0.24359512809743805,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.121287128712871e-06,
+      "loss": 1.9225,
+      "step": 406
+    },
+    {
+      "epoch": 0.24419511609767805,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.1188118811881187e-06,
+      "loss": 1.8012,
+      "step": 407
+    },
+    {
+      "epoch": 0.24479510409791805,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.116336633663366e-06,
+      "loss": 1.8264,
+      "step": 408
+    },
+    {
+      "epoch": 0.24539509209815805,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.113861386138614e-06,
+      "loss": 1.9196,
+      "step": 409
+    },
+    {
+      "epoch": 0.24599508009839804,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.111386138613861e-06,
+      "loss": 1.7687,
+      "step": 410
+    },
+    {
+      "epoch": 0.24659506809863802,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.108910891089109e-06,
+      "loss": 1.8449,
+      "step": 411
+    },
+    {
+      "epoch": 0.247195056098878,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.1064356435643563e-06,
+      "loss": 1.9169,
+      "step": 412
+    },
+    {
+      "epoch": 0.247795044099118,
+      "grad_norm": 1.4453125,
+      "learning_rate": 3.103960396039604e-06,
+      "loss": 1.83,
+      "step": 413
+    },
+    {
+      "epoch": 0.248395032099358,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.1014851485148514e-06,
+      "loss": 1.8564,
+      "step": 414
+    },
+    {
+      "epoch": 0.248995020099598,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.099009900990099e-06,
+      "loss": 1.8824,
+      "step": 415
+    },
+    {
+      "epoch": 0.249595008099838,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.096534653465346e-06,
+      "loss": 1.7244,
+      "step": 416
+    },
+    {
+      "epoch": 0.250194996100078,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.094059405940594e-06,
+      "loss": 1.7585,
+      "step": 417
+    },
+    {
+      "epoch": 0.250794984100318,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.091584158415841e-06,
+      "loss": 1.8895,
+      "step": 418
+    },
+    {
+      "epoch": 0.251394972100558,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.089108910891089e-06,
+      "loss": 1.8884,
+      "step": 419
+    },
+    {
+      "epoch": 0.25199496010079797,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.0866336633663363e-06,
+      "loss": 1.8346,
+      "step": 420
+    },
+    {
+      "epoch": 0.252594948101038,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.084158415841584e-06,
+      "loss": 1.8132,
+      "step": 421
+    },
+    {
+      "epoch": 0.25319493610127797,
+      "grad_norm": 1.125,
+      "learning_rate": 3.0816831683168314e-06,
+      "loss": 1.7638,
+      "step": 422
+    },
+    {
+      "epoch": 0.253794924101518,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.079207920792079e-06,
+      "loss": 1.8337,
+      "step": 423
+    },
+    {
+      "epoch": 0.25439491210175796,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.0767326732673265e-06,
+      "loss": 1.8597,
+      "step": 424
+    },
+    {
+      "epoch": 0.25499490010199793,
+      "grad_norm": 4.25,
+      "learning_rate": 3.0742574257425743e-06,
+      "loss": 1.7829,
+      "step": 425
+    },
+    {
+      "epoch": 0.25559488810223796,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.0717821782178216e-06,
+      "loss": 1.8581,
+      "step": 426
+    },
+    {
+      "epoch": 0.25619487610247793,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.0693069306930694e-06,
+      "loss": 1.8007,
+      "step": 427
+    },
+    {
+      "epoch": 0.25679486410271796,
+      "grad_norm": 1.359375,
+      "learning_rate": 3.0668316831683167e-06,
+      "loss": 1.7999,
+      "step": 428
+    },
+    {
+      "epoch": 0.2573948521029579,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.064356435643564e-06,
+      "loss": 1.9848,
+      "step": 429
+    },
+    {
+      "epoch": 0.25799484010319795,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.0618811881188114e-06,
+      "loss": 1.7019,
+      "step": 430
+    },
+    {
+      "epoch": 0.2585948281034379,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.059405940594059e-06,
+      "loss": 1.8219,
+      "step": 431
+    },
+    {
+      "epoch": 0.25919481610367795,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.0569306930693065e-06,
+      "loss": 1.8192,
+      "step": 432
+    },
+    {
+      "epoch": 0.2597948041039179,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.0544554455445543e-06,
+      "loss": 1.9334,
+      "step": 433
+    },
+    {
+      "epoch": 0.2603947921041579,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.0519801980198016e-06,
+      "loss": 1.7363,
+      "step": 434
+    },
+    {
+      "epoch": 0.2609947801043979,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.0495049504950494e-06,
+      "loss": 1.8116,
+      "step": 435
+    },
+    {
+      "epoch": 0.2615947681046379,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.0470297029702967e-06,
+      "loss": 1.849,
+      "step": 436
+    },
+    {
+      "epoch": 0.2621947561048779,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.0445544554455445e-06,
+      "loss": 1.8584,
+      "step": 437
+    },
+    {
+      "epoch": 0.2627947441051179,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.042079207920792e-06,
+      "loss": 1.7644,
+      "step": 438
+    },
+    {
+      "epoch": 0.2633947321053579,
+      "grad_norm": 1.34375,
+      "learning_rate": 3.0396039603960396e-06,
+      "loss": 1.71,
+      "step": 439
+    },
+    {
+      "epoch": 0.2639947201055979,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.037128712871287e-06,
+      "loss": 1.7796,
+      "step": 440
+    },
+    {
+      "epoch": 0.2645947081058379,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.0346534653465347e-06,
+      "loss": 1.8207,
+      "step": 441
+    },
+    {
+      "epoch": 0.2651946961060779,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.0321782178217817e-06,
+      "loss": 1.8873,
+      "step": 442
+    },
+    {
+      "epoch": 0.26579468410631785,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.0297029702970294e-06,
+      "loss": 1.9306,
+      "step": 443
+    },
+    {
+      "epoch": 0.2663946721065579,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.0272277227722768e-06,
+      "loss": 1.8683,
+      "step": 444
+    },
+    {
+      "epoch": 0.26699466010679784,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.0247524752475245e-06,
+      "loss": 1.8457,
+      "step": 445
+    },
+    {
+      "epoch": 0.26759464810703787,
+      "grad_norm": 1.40625,
+      "learning_rate": 3.022277227722772e-06,
+      "loss": 1.7025,
+      "step": 446
+    },
+    {
+      "epoch": 0.26819463610727784,
+      "grad_norm": 1.7421875,
+      "learning_rate": 3.0198019801980196e-06,
+      "loss": 1.7783,
+      "step": 447
+    },
+    {
+      "epoch": 0.26879462410751787,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.017326732673267e-06,
+      "loss": 1.9892,
+      "step": 448
+    },
+    {
+      "epoch": 0.26939461210775784,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.0148514851485147e-06,
+      "loss": 1.7737,
+      "step": 449
+    },
+    {
+      "epoch": 0.26999460010799786,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.012376237623762e-06,
+      "loss": 1.8656,
+      "step": 450
+    },
+    {
+      "epoch": 0.26999460010799786,
+      "eval_loss": 2.033843994140625,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 67.9747,
+      "eval_samples_per_second": 147.114,
+      "eval_steps_per_second": 24.524,
+      "step": 450
+    },
+    {
+      "epoch": 0.27059458810823783,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.00990099009901e-06,
+      "loss": 1.7738,
+      "step": 451
+    },
+    {
+      "epoch": 0.2711945761084778,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.007425742574257e-06,
+      "loss": 1.8191,
+      "step": 452
+    },
+    {
+      "epoch": 0.27179456410871783,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.004950495049505e-06,
+      "loss": 1.7908,
+      "step": 453
+    },
+    {
+      "epoch": 0.2723945521089578,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.0024752475247523e-06,
+      "loss": 1.7694,
+      "step": 454
+    },
+    {
+      "epoch": 0.2729945401091978,
+      "grad_norm": 1.40625,
+      "learning_rate": 3e-06,
+      "loss": 1.8002,
+      "step": 455
+    },
+    {
+      "epoch": 0.2735945281094378,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.9975247524752474e-06,
+      "loss": 1.8376,
+      "step": 456
+    },
+    {
+      "epoch": 0.2741945161096778,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.9950495049504948e-06,
+      "loss": 1.8418,
+      "step": 457
+    },
+    {
+      "epoch": 0.2747945041099178,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.9925742574257425e-06,
+      "loss": 1.9308,
+      "step": 458
+    },
+    {
+      "epoch": 0.2753944921101578,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.99009900990099e-06,
+      "loss": 1.7999,
+      "step": 459
+    },
+    {
+      "epoch": 0.2759944801103978,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.9876237623762376e-06,
+      "loss": 1.8369,
+      "step": 460
+    },
+    {
+      "epoch": 0.2765944681106378,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.985148514851485e-06,
+      "loss": 1.8621,
+      "step": 461
+    },
+    {
+      "epoch": 0.2771944561108778,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.9826732673267327e-06,
+      "loss": 1.7909,
+      "step": 462
+    },
+    {
+      "epoch": 0.27779444411111776,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.98019801980198e-06,
+      "loss": 1.7593,
+      "step": 463
+    },
+    {
+      "epoch": 0.2783944321113578,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.977722772277228e-06,
+      "loss": 1.8737,
+      "step": 464
+    },
+    {
+      "epoch": 0.27899442011159775,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.975247524752475e-06,
+      "loss": 1.7006,
+      "step": 465
+    },
+    {
+      "epoch": 0.2795944081118378,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.972772277227723e-06,
+      "loss": 1.9047,
+      "step": 466
+    },
+    {
+      "epoch": 0.28019439611207775,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.9702970297029703e-06,
+      "loss": 1.7895,
+      "step": 467
+    },
+    {
+      "epoch": 0.2807943841123178,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.967821782178218e-06,
+      "loss": 1.7947,
+      "step": 468
+    },
+    {
+      "epoch": 0.28139437211255774,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.965346534653465e-06,
+      "loss": 1.7682,
+      "step": 469
+    },
+    {
+      "epoch": 0.28199436011279777,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.9628712871287128e-06,
+      "loss": 1.8296,
+      "step": 470
+    },
+    {
+      "epoch": 0.28259434811303774,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.96039603960396e-06,
+      "loss": 1.9427,
+      "step": 471
+    },
+    {
+      "epoch": 0.2831943361132777,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.957920792079208e-06,
+      "loss": 1.9328,
+      "step": 472
+    },
+    {
+      "epoch": 0.28379432411351774,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.9554455445544552e-06,
+      "loss": 1.8043,
+      "step": 473
+    },
+    {
+      "epoch": 0.2843943121137577,
+      "grad_norm": 2.375,
+      "learning_rate": 2.952970297029703e-06,
+      "loss": 1.7856,
+      "step": 474
+    },
+    {
+      "epoch": 0.28499430011399773,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.9504950495049503e-06,
+      "loss": 1.7967,
+      "step": 475
+    },
+    {
+      "epoch": 0.2855942881142377,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.948019801980198e-06,
+      "loss": 1.8577,
+      "step": 476
+    },
+    {
+      "epoch": 0.28619427611447773,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.9455445544554454e-06,
+      "loss": 1.8487,
+      "step": 477
+    },
+    {
+      "epoch": 0.2867942641147177,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.943069306930693e-06,
+      "loss": 1.9896,
+      "step": 478
+    },
+    {
+      "epoch": 0.2873942521149577,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.9405940594059405e-06,
+      "loss": 1.9078,
+      "step": 479
+    },
+    {
+      "epoch": 0.2879942401151977,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.9381188118811883e-06,
+      "loss": 1.8419,
+      "step": 480
+    },
+    {
+      "epoch": 0.28859422811543767,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.9356435643564357e-06,
+      "loss": 1.8428,
+      "step": 481
+    },
+    {
+      "epoch": 0.2891942161156777,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.933168316831683e-06,
+      "loss": 1.9351,
+      "step": 482
+    },
+    {
+      "epoch": 0.28979420411591766,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.9306930693069303e-06,
+      "loss": 1.8918,
+      "step": 483
+    },
+    {
+      "epoch": 0.2903941921161577,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.928217821782178e-06,
+      "loss": 2.0012,
+      "step": 484
+    },
+    {
+      "epoch": 0.29099418011639766,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.9257425742574254e-06,
+      "loss": 1.7536,
+      "step": 485
+    },
+    {
+      "epoch": 0.2915941681166377,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.9232673267326732e-06,
+      "loss": 1.8447,
+      "step": 486
+    },
+    {
+      "epoch": 0.29219415611687766,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.9207920792079206e-06,
+      "loss": 1.7318,
+      "step": 487
+    },
+    {
+      "epoch": 0.2927941441171177,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.9183168316831683e-06,
+      "loss": 1.8265,
+      "step": 488
+    },
+    {
+      "epoch": 0.29339413211735765,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.9158415841584157e-06,
+      "loss": 1.7215,
+      "step": 489
+    },
+    {
+      "epoch": 0.2939941201175976,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.9133663366336634e-06,
+      "loss": 1.8658,
+      "step": 490
+    },
+    {
+      "epoch": 0.29459410811783765,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.9108910891089108e-06,
+      "loss": 1.8764,
+      "step": 491
+    },
+    {
+      "epoch": 0.2951940961180776,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.9084158415841585e-06,
+      "loss": 1.9107,
+      "step": 492
+    },
+    {
+      "epoch": 0.29579408411831765,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.905940594059406e-06,
+      "loss": 1.8486,
+      "step": 493
+    },
+    {
+      "epoch": 0.2963940721185576,
+      "grad_norm": 1.2890625,
+      "learning_rate": 2.9034653465346537e-06,
+      "loss": 1.7087,
+      "step": 494
+    },
+    {
+      "epoch": 0.29699406011879764,
+      "grad_norm": 1.25,
+      "learning_rate": 2.9009900990099006e-06,
+      "loss": 1.8336,
+      "step": 495
+    },
+    {
+      "epoch": 0.2975940481190376,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.8985148514851483e-06,
+      "loss": 1.9546,
+      "step": 496
+    },
+    {
+      "epoch": 0.29819403611927764,
+      "grad_norm": 1.2890625,
+      "learning_rate": 2.8960396039603957e-06,
+      "loss": 1.8021,
+      "step": 497
+    },
+    {
+      "epoch": 0.2987940241195176,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.8935643564356434e-06,
+      "loss": 1.8557,
+      "step": 498
+    },
+    {
+      "epoch": 0.2993940121197576,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.891089108910891e-06,
+      "loss": 1.8676,
+      "step": 499
+    },
+    {
+      "epoch": 0.2999940001199976,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.8886138613861386e-06,
+      "loss": 1.8587,
+      "step": 500
+    },
+    {
+      "epoch": 0.2999940001199976,
+      "eval_loss": 2.0334386825561523,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.1108,
+      "eval_samples_per_second": 153.584,
+      "eval_steps_per_second": 25.603,
+      "step": 500
+    },
+    {
+      "epoch": 0.3005939881202376,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2.886138613861386e-06,
+      "loss": 1.828,
+      "step": 501
+    },
+    {
+      "epoch": 0.3011939761204776,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.8836633663366337e-06,
+      "loss": 1.8682,
+      "step": 502
+    },
+    {
+      "epoch": 0.3017939641207176,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.881188118811881e-06,
+      "loss": 1.7229,
+      "step": 503
+    },
+    {
+      "epoch": 0.3023939521209576,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8787128712871288e-06,
+      "loss": 1.9047,
+      "step": 504
+    },
+    {
+      "epoch": 0.30299394012119757,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.876237623762376e-06,
+      "loss": 1.8765,
+      "step": 505
+    },
+    {
+      "epoch": 0.3035939281214376,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.873762376237624e-06,
+      "loss": 1.8074,
+      "step": 506
+    },
+    {
+      "epoch": 0.30419391612167757,
+      "grad_norm": 1.46875,
+      "learning_rate": 2.8712871287128712e-06,
+      "loss": 1.8522,
+      "step": 507
+    },
+    {
+      "epoch": 0.30479390412191754,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.868811881188119e-06,
+      "loss": 1.7335,
+      "step": 508
+    },
+    {
+      "epoch": 0.30539389212215756,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.866336633663366e-06,
+      "loss": 1.8103,
+      "step": 509
+    },
+    {
+      "epoch": 0.30599388012239753,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.8638613861386137e-06,
+      "loss": 1.7559,
+      "step": 510
+    },
+    {
+      "epoch": 0.30659386812263756,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.861386138613861e-06,
+      "loss": 1.7131,
+      "step": 511
+    },
+    {
+      "epoch": 0.30719385612287753,
+      "grad_norm": 1.375,
+      "learning_rate": 2.858910891089109e-06,
+      "loss": 1.889,
+      "step": 512
+    },
+    {
+      "epoch": 0.30779384412311755,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.856435643564356e-06,
+      "loss": 1.843,
+      "step": 513
+    },
+    {
+      "epoch": 0.3083938321233575,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.853960396039604e-06,
+      "loss": 1.7522,
+      "step": 514
+    },
+    {
+      "epoch": 0.30899382012359755,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.8514851485148512e-06,
+      "loss": 1.8443,
+      "step": 515
+    },
+    {
+      "epoch": 0.3095938081238375,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.849009900990099e-06,
+      "loss": 1.7471,
+      "step": 516
+    },
+    {
+      "epoch": 0.3101937961240775,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.8465346534653464e-06,
+      "loss": 1.8667,
+      "step": 517
+    },
+    {
+      "epoch": 0.3107937841243175,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.844059405940594e-06,
+      "loss": 1.7233,
+      "step": 518
+    },
+    {
+      "epoch": 0.3113937721245575,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.8415841584158415e-06,
+      "loss": 1.8619,
+      "step": 519
+    },
+    {
+      "epoch": 0.3119937601247975,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.8391089108910892e-06,
+      "loss": 1.8094,
+      "step": 520
+    },
+    {
+      "epoch": 0.3125937481250375,
+      "grad_norm": 1.3515625,
+      "learning_rate": 2.8366336633663366e-06,
+      "loss": 1.8636,
+      "step": 521
+    },
+    {
+      "epoch": 0.3131937361252775,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.834158415841584e-06,
+      "loss": 1.8544,
+      "step": 522
+    },
+    {
+      "epoch": 0.3137937241255175,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.8316831683168313e-06,
+      "loss": 1.7636,
+      "step": 523
+    },
+    {
+      "epoch": 0.3143937121257575,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.829207920792079e-06,
+      "loss": 1.8123,
+      "step": 524
+    },
+    {
+      "epoch": 0.3149937001259975,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.8267326732673264e-06,
+      "loss": 1.7643,
+      "step": 525
+    },
+    {
+      "epoch": 0.31559368812623745,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.824257425742574e-06,
+      "loss": 1.8393,
+      "step": 526
+    },
+    {
+      "epoch": 0.3161936761264775,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.8217821782178215e-06,
+      "loss": 1.7629,
+      "step": 527
+    },
+    {
+      "epoch": 0.31679366412671744,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.8193069306930692e-06,
+      "loss": 1.8098,
+      "step": 528
+    },
+    {
+      "epoch": 0.31739365212695747,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.8168316831683166e-06,
+      "loss": 1.7945,
+      "step": 529
+    },
+    {
+      "epoch": 0.31799364012719744,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.8143564356435644e-06,
+      "loss": 1.7641,
+      "step": 530
+    },
+    {
+      "epoch": 0.31859362812743747,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.8118811881188117e-06,
+      "loss": 1.9744,
+      "step": 531
+    },
+    {
+      "epoch": 0.31919361612767744,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.8094059405940595e-06,
+      "loss": 1.7955,
+      "step": 532
+    },
+    {
+      "epoch": 0.31979360412791746,
+      "grad_norm": 1.40625,
+      "learning_rate": 2.806930693069307e-06,
+      "loss": 1.8048,
+      "step": 533
+    },
+    {
+      "epoch": 0.32039359212815743,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.8044554455445546e-06,
+      "loss": 1.8906,
+      "step": 534
+    },
+    {
+      "epoch": 0.3209935801283974,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.8019801980198015e-06,
+      "loss": 1.8028,
+      "step": 535
+    },
+    {
+      "epoch": 0.32159356812863743,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.7995049504950493e-06,
+      "loss": 1.7173,
+      "step": 536
+    },
+    {
+      "epoch": 0.3221935561288774,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.7970297029702966e-06,
+      "loss": 1.8689,
+      "step": 537
+    },
+    {
+      "epoch": 0.3227935441291174,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.7945544554455444e-06,
+      "loss": 1.7717,
+      "step": 538
+    },
+    {
+      "epoch": 0.3233935321293574,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.7920792079207917e-06,
+      "loss": 1.9151,
+      "step": 539
+    },
+    {
+      "epoch": 0.3239935201295974,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.7896039603960395e-06,
+      "loss": 1.9621,
+      "step": 540
+    },
+    {
+      "epoch": 0.3245935081298374,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.787128712871287e-06,
+      "loss": 1.801,
+      "step": 541
+    },
+    {
+      "epoch": 0.3251934961300774,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.7846534653465346e-06,
+      "loss": 1.9634,
+      "step": 542
+    },
+    {
+      "epoch": 0.3257934841303174,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.782178217821782e-06,
+      "loss": 1.7282,
+      "step": 543
+    },
+    {
+      "epoch": 0.3263934721305574,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.7797029702970297e-06,
+      "loss": 1.6894,
+      "step": 544
+    },
+    {
+      "epoch": 0.3269934601307974,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.777227722772277e-06,
+      "loss": 1.7942,
+      "step": 545
+    },
+    {
+      "epoch": 0.32759344813103736,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.774752475247525e-06,
+      "loss": 1.7971,
+      "step": 546
+    },
+    {
+      "epoch": 0.3281934361312774,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.772277227722772e-06,
+      "loss": 1.7986,
+      "step": 547
+    },
+    {
+      "epoch": 0.32879342413151735,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.7698019801980195e-06,
+      "loss": 2.0276,
+      "step": 548
+    },
+    {
+      "epoch": 0.3293934121317574,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.767326732673267e-06,
+      "loss": 1.8949,
+      "step": 549
+    },
+    {
+      "epoch": 0.32999340013199735,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.7648514851485146e-06,
+      "loss": 1.8993,
+      "step": 550
+    },
+    {
+      "epoch": 0.32999340013199735,
+      "eval_loss": 2.0332491397857666,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.0973,
+      "eval_samples_per_second": 153.616,
+      "eval_steps_per_second": 25.608,
+      "step": 550
+    },
+    {
+      "epoch": 0.3305933881322374,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.762376237623762e-06,
+      "loss": 1.7827,
+      "step": 551
+    },
+    {
+      "epoch": 0.33119337613247735,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.7599009900990097e-06,
+      "loss": 1.754,
+      "step": 552
+    },
+    {
+      "epoch": 0.33179336413271737,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.757425742574257e-06,
+      "loss": 1.7121,
+      "step": 553
+    },
+    {
+      "epoch": 0.33239335213295734,
+      "grad_norm": 1.703125,
+      "learning_rate": 2.754950495049505e-06,
+      "loss": 1.8265,
+      "step": 554
+    },
+    {
+      "epoch": 0.3329933401331973,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.752475247524752e-06,
+      "loss": 1.8452,
+      "step": 555
+    },
+    {
+      "epoch": 0.33359332813343734,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.75e-06,
+      "loss": 1.9355,
+      "step": 556
+    },
+    {
+      "epoch": 0.3341933161336773,
+      "grad_norm": 1.5859375,
+      "learning_rate": 2.7475247524752477e-06,
+      "loss": 1.7684,
+      "step": 557
+    },
+    {
+      "epoch": 0.33479330413391734,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.745049504950495e-06,
+      "loss": 1.8229,
+      "step": 558
+    },
+    {
+      "epoch": 0.3353932921341573,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.742574257425743e-06,
+      "loss": 1.9012,
+      "step": 559
+    },
+    {
+      "epoch": 0.33599328013439733,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.74009900990099e-06,
+      "loss": 1.8624,
+      "step": 560
+    },
+    {
+      "epoch": 0.3365932681346373,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.737623762376238e-06,
+      "loss": 1.8385,
+      "step": 561
+    },
+    {
+      "epoch": 0.33719325613487733,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.735148514851485e-06,
+      "loss": 1.9174,
+      "step": 562
+    },
+    {
+      "epoch": 0.3377932441351173,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.7326732673267326e-06,
+      "loss": 1.6744,
+      "step": 563
+    },
+    {
+      "epoch": 0.33839323213535727,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.73019801980198e-06,
+      "loss": 1.9218,
+      "step": 564
+    },
+    {
+      "epoch": 0.3389932201355973,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.7277227722772277e-06,
+      "loss": 1.9251,
+      "step": 565
+    },
+    {
+      "epoch": 0.33959320813583727,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.725247524752475e-06,
+      "loss": 2.0032,
+      "step": 566
+    },
+    {
+      "epoch": 0.3401931961360773,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.722772277227723e-06,
+      "loss": 1.7809,
+      "step": 567
+    },
+    {
+      "epoch": 0.34079318413631726,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.72029702970297e-06,
+      "loss": 1.7717,
+      "step": 568
+    },
+    {
+      "epoch": 0.3413931721365573,
+      "grad_norm": 1.25,
+      "learning_rate": 2.717821782178218e-06,
+      "loss": 1.881,
+      "step": 569
+    },
+    {
+      "epoch": 0.34199316013679726,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.7153465346534653e-06,
+      "loss": 1.8142,
+      "step": 570
+    },
+    {
+      "epoch": 0.3425931481370373,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.712871287128713e-06,
+      "loss": 1.9001,
+      "step": 571
+    },
+    {
+      "epoch": 0.34319313613727725,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.7103960396039604e-06,
+      "loss": 1.7508,
+      "step": 572
+    },
+    {
+      "epoch": 0.3437931241375172,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.707920792079208e-06,
+      "loss": 2.0335,
+      "step": 573
+    },
+    {
+      "epoch": 0.34439311213775725,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.7054455445544555e-06,
+      "loss": 1.9974,
+      "step": 574
+    },
+    {
+      "epoch": 0.3449931001379972,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.702970297029703e-06,
+      "loss": 1.9025,
+      "step": 575
+    },
+    {
+      "epoch": 0.34559308813823725,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.70049504950495e-06,
+      "loss": 1.7668,
+      "step": 576
+    },
+    {
+      "epoch": 0.3461930761384772,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.698019801980198e-06,
+      "loss": 1.8825,
+      "step": 577
+    },
+    {
+      "epoch": 0.34679306413871724,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.6955445544554453e-06,
+      "loss": 1.9186,
+      "step": 578
+    },
+    {
+      "epoch": 0.3473930521389572,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.693069306930693e-06,
+      "loss": 1.8675,
+      "step": 579
+    },
+    {
+      "epoch": 0.34799304013919724,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.6905940594059404e-06,
+      "loss": 1.9197,
+      "step": 580
+    },
+    {
+      "epoch": 0.3485930281394372,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.688118811881188e-06,
+      "loss": 1.7689,
+      "step": 581
+    },
+    {
+      "epoch": 0.3491930161396772,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.6856435643564355e-06,
+      "loss": 1.7074,
+      "step": 582
+    },
+    {
+      "epoch": 0.3497930041399172,
+      "grad_norm": 1.546875,
+      "learning_rate": 2.6831683168316833e-06,
+      "loss": 1.8839,
+      "step": 583
+    },
+    {
+      "epoch": 0.3503929921401572,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.6806930693069306e-06,
+      "loss": 1.7625,
+      "step": 584
+    },
+    {
+      "epoch": 0.3509929801403972,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.6782178217821784e-06,
+      "loss": 1.767,
+      "step": 585
+    },
+    {
+      "epoch": 0.3515929681406372,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.6757425742574257e-06,
+      "loss": 1.9642,
+      "step": 586
+    },
+    {
+      "epoch": 0.3521929561408772,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.6732673267326735e-06,
+      "loss": 1.9743,
+      "step": 587
+    },
+    {
+      "epoch": 0.35279294414111717,
+      "grad_norm": 1.3515625,
+      "learning_rate": 2.6707920792079204e-06,
+      "loss": 1.9189,
+      "step": 588
+    },
+    {
+      "epoch": 0.3533929321413572,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.668316831683168e-06,
+      "loss": 1.793,
+      "step": 589
+    },
+    {
+      "epoch": 0.35399292014159717,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.6658415841584155e-06,
+      "loss": 1.8837,
+      "step": 590
+    },
+    {
+      "epoch": 0.35459290814183714,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.6633663366336633e-06,
+      "loss": 1.8041,
+      "step": 591
+    },
+    {
+      "epoch": 0.35519289614207716,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.6608910891089106e-06,
+      "loss": 1.797,
+      "step": 592
+    },
+    {
+      "epoch": 0.35579288414231713,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.6584158415841584e-06,
+      "loss": 1.6775,
+      "step": 593
+    },
+    {
+      "epoch": 0.35639287214255716,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.6559405940594057e-06,
+      "loss": 1.8262,
+      "step": 594
+    },
+    {
+      "epoch": 0.35699286014279713,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.6534653465346535e-06,
+      "loss": 1.8204,
+      "step": 595
+    },
+    {
+      "epoch": 0.35759284814303716,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.650990099009901e-06,
+      "loss": 1.8847,
+      "step": 596
+    },
+    {
+      "epoch": 0.3581928361432771,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.6485148514851486e-06,
+      "loss": 1.8749,
+      "step": 597
+    },
+    {
+      "epoch": 0.35879282414351715,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.646039603960396e-06,
+      "loss": 1.816,
+      "step": 598
+    },
+    {
+      "epoch": 0.3593928121437571,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.6435643564356437e-06,
+      "loss": 1.8531,
+      "step": 599
+    },
+    {
+      "epoch": 0.3599928001439971,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.641089108910891e-06,
+      "loss": 1.8029,
+      "step": 600
+    },
+    {
+      "epoch": 0.3599928001439971,
+      "eval_loss": 2.0328757762908936,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 67.4254,
+      "eval_samples_per_second": 148.312,
+      "eval_steps_per_second": 24.724,
+      "step": 600
+    },
+    {
+      "epoch": 0.3605927881442371,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.6386138613861384e-06,
+      "loss": 1.8685,
+      "step": 601
+    },
+    {
+      "epoch": 0.3611927761444771,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.6361386138613858e-06,
+      "loss": 1.8958,
+      "step": 602
+    },
+    {
+      "epoch": 0.3617927641447171,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.6336633663366335e-06,
+      "loss": 1.7194,
+      "step": 603
+    },
+    {
+      "epoch": 0.3623927521449571,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.631188118811881e-06,
+      "loss": 1.9711,
+      "step": 604
+    },
+    {
+      "epoch": 0.3629927401451971,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.6287128712871286e-06,
+      "loss": 1.7863,
+      "step": 605
+    },
+    {
+      "epoch": 0.3635927281454371,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.626237623762376e-06,
+      "loss": 1.8361,
+      "step": 606
+    },
+    {
+      "epoch": 0.3641927161456771,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.6237623762376237e-06,
+      "loss": 1.9021,
+      "step": 607
+    },
+    {
+      "epoch": 0.3647927041459171,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.621287128712871e-06,
+      "loss": 1.9613,
+      "step": 608
+    },
+    {
+      "epoch": 0.36539269214615705,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.618811881188119e-06,
+      "loss": 1.8389,
+      "step": 609
+    },
+    {
+      "epoch": 0.3659926801463971,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.616336633663366e-06,
+      "loss": 1.8752,
+      "step": 610
+    },
+    {
+      "epoch": 0.36659266814663705,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.613861386138614e-06,
+      "loss": 1.8041,
+      "step": 611
+    },
+    {
+      "epoch": 0.36719265614687707,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.6113861386138613e-06,
+      "loss": 1.8281,
+      "step": 612
+    },
+    {
+      "epoch": 0.36779264414711704,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.608910891089109e-06,
+      "loss": 1.7512,
+      "step": 613
+    },
+    {
+      "epoch": 0.36839263214735707,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.606435643564356e-06,
+      "loss": 1.9474,
+      "step": 614
+    },
+    {
+      "epoch": 0.36899262014759704,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.6039603960396038e-06,
+      "loss": 1.8229,
+      "step": 615
+    },
+    {
+      "epoch": 0.36959260814783707,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.601485148514851e-06,
+      "loss": 1.6716,
+      "step": 616
+    },
+    {
+      "epoch": 0.37019259614807704,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.599009900990099e-06,
+      "loss": 1.958,
+      "step": 617
+    },
+    {
+      "epoch": 0.370792584148317,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.596534653465346e-06,
+      "loss": 1.886,
+      "step": 618
+    },
+    {
+      "epoch": 0.37139257214855703,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.594059405940594e-06,
+      "loss": 1.904,
+      "step": 619
+    },
+    {
+      "epoch": 0.371992560148797,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.5915841584158413e-06,
+      "loss": 1.8177,
+      "step": 620
+    },
+    {
+      "epoch": 0.37259254814903703,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.589108910891089e-06,
+      "loss": 1.7927,
+      "step": 621
+    },
+    {
+      "epoch": 0.373192536149277,
+      "grad_norm": 1.2890625,
+      "learning_rate": 2.5866336633663364e-06,
+      "loss": 1.8946,
+      "step": 622
+    },
+    {
+      "epoch": 0.373792524149517,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.584158415841584e-06,
+      "loss": 1.7751,
+      "step": 623
+    },
+    {
+      "epoch": 0.374392512149757,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.5816831683168315e-06,
+      "loss": 1.9221,
+      "step": 624
+    },
+    {
+      "epoch": 0.374992500149997,
+      "grad_norm": 1.3515625,
+      "learning_rate": 2.5792079207920793e-06,
+      "loss": 1.9133,
+      "step": 625
+    },
+    {
+      "epoch": 0.375592488150237,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.5767326732673266e-06,
+      "loss": 1.8273,
+      "step": 626
+    },
+    {
+      "epoch": 0.376192476150477,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.5742574257425744e-06,
+      "loss": 1.7544,
+      "step": 627
+    },
+    {
+      "epoch": 0.376792464150717,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.5717821782178213e-06,
+      "loss": 1.7994,
+      "step": 628
+    },
+    {
+      "epoch": 0.37739245215095696,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.569306930693069e-06,
+      "loss": 1.9355,
+      "step": 629
+    },
+    {
+      "epoch": 0.377992440151197,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.5668316831683164e-06,
+      "loss": 1.7472,
+      "step": 630
+    },
+    {
+      "epoch": 0.37859242815143695,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.564356435643564e-06,
+      "loss": 1.9562,
+      "step": 631
+    },
+    {
+      "epoch": 0.379192416151677,
+      "grad_norm": 1.4765625,
+      "learning_rate": 2.5618811881188115e-06,
+      "loss": 1.9807,
+      "step": 632
+    },
+    {
+      "epoch": 0.37979240415191695,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.5594059405940593e-06,
+      "loss": 1.9289,
+      "step": 633
+    },
+    {
+      "epoch": 0.380392392152157,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.5569306930693067e-06,
+      "loss": 1.9202,
+      "step": 634
+    },
+    {
+      "epoch": 0.38099238015239695,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.5544554455445544e-06,
+      "loss": 1.8881,
+      "step": 635
+    },
+    {
+      "epoch": 0.381592368152637,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.5519801980198018e-06,
+      "loss": 1.8555,
+      "step": 636
+    },
+    {
+      "epoch": 0.38219235615287694,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.5495049504950495e-06,
+      "loss": 1.9314,
+      "step": 637
+    },
+    {
+      "epoch": 0.3827923441531169,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.547029702970297e-06,
+      "loss": 1.7896,
+      "step": 638
+    },
+    {
+      "epoch": 0.38339233215335694,
+      "grad_norm": 1.4375,
+      "learning_rate": 2.5445544554455446e-06,
+      "loss": 1.8339,
+      "step": 639
+    },
+    {
+      "epoch": 0.3839923201535969,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.542079207920792e-06,
+      "loss": 1.9142,
+      "step": 640
+    },
+    {
+      "epoch": 0.38459230815383694,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.5396039603960393e-06,
+      "loss": 1.9154,
+      "step": 641
+    },
+    {
+      "epoch": 0.3851922961540769,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.5371287128712867e-06,
+      "loss": 1.8165,
+      "step": 642
+    },
+    {
+      "epoch": 0.38579228415431693,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.5346534653465344e-06,
+      "loss": 1.6729,
+      "step": 643
+    },
+    {
+      "epoch": 0.3863922721545569,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.5321782178217818e-06,
+      "loss": 1.8441,
+      "step": 644
+    },
+    {
+      "epoch": 0.38699226015479693,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.5297029702970295e-06,
+      "loss": 1.7124,
+      "step": 645
+    },
+    {
+      "epoch": 0.3875922481550369,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.527227722772277e-06,
+      "loss": 1.8526,
+      "step": 646
+    },
+    {
+      "epoch": 0.38819223615527687,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.5247524752475247e-06,
+      "loss": 1.7846,
+      "step": 647
+    },
+    {
+      "epoch": 0.3887922241555169,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.522277227722772e-06,
+      "loss": 1.8099,
+      "step": 648
+    },
+    {
+      "epoch": 0.38939221215575687,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.5198019801980198e-06,
+      "loss": 1.8258,
+      "step": 649
+    },
+    {
+      "epoch": 0.3899922001559969,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.517326732673267e-06,
+      "loss": 1.7954,
+      "step": 650
+    },
+    {
+      "epoch": 0.3899922001559969,
+      "eval_loss": 2.032578468322754,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.0425,
+      "eval_samples_per_second": 153.746,
+      "eval_steps_per_second": 25.629,
+      "step": 650
+    },
+    {
+      "epoch": 0.39059218815623686,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.514851485148515e-06,
+      "loss": 1.8582,
+      "step": 651
+    },
+    {
+      "epoch": 0.3911921761564769,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.5123762376237622e-06,
+      "loss": 1.7973,
+      "step": 652
+    },
+    {
+      "epoch": 0.39179216415671686,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.50990099009901e-06,
+      "loss": 1.8664,
+      "step": 653
+    },
+    {
+      "epoch": 0.3923921521569569,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.507425742574257e-06,
+      "loss": 1.8547,
+      "step": 654
+    },
+    {
+      "epoch": 0.39299214015719686,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.5049504950495047e-06,
+      "loss": 1.7987,
+      "step": 655
+    },
+    {
+      "epoch": 0.3935921281574368,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.502475247524752e-06,
+      "loss": 1.7597,
+      "step": 656
+    },
+    {
+      "epoch": 0.39419211615767685,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.4999999999999998e-06,
+      "loss": 1.8468,
+      "step": 657
+    },
+    {
+      "epoch": 0.3947921041579168,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.4975247524752475e-06,
+      "loss": 1.7945,
+      "step": 658
+    },
+    {
+      "epoch": 0.39539209215815685,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.495049504950495e-06,
+      "loss": 1.7316,
+      "step": 659
+    },
+    {
+      "epoch": 0.3959920801583968,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.4925742574257427e-06,
+      "loss": 1.7246,
+      "step": 660
+    },
+    {
+      "epoch": 0.39659206815863685,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.49009900990099e-06,
+      "loss": 1.7835,
+      "step": 661
+    },
+    {
+      "epoch": 0.3971920561588768,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.4876237623762378e-06,
+      "loss": 1.8527,
+      "step": 662
+    },
+    {
+      "epoch": 0.39779204415911684,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.485148514851485e-06,
+      "loss": 1.8803,
+      "step": 663
+    },
+    {
+      "epoch": 0.3983920321593568,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.482673267326733e-06,
+      "loss": 1.8751,
+      "step": 664
+    },
+    {
+      "epoch": 0.3989920201595968,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.4801980198019802e-06,
+      "loss": 1.9245,
+      "step": 665
+    },
+    {
+      "epoch": 0.3995920081598368,
+      "grad_norm": 1.25,
+      "learning_rate": 2.477722772277228e-06,
+      "loss": 1.7848,
+      "step": 666
+    },
+    {
+      "epoch": 0.4001919961600768,
+      "grad_norm": 1.9765625,
+      "learning_rate": 2.475247524752475e-06,
+      "loss": 1.7866,
+      "step": 667
+    },
+    {
+      "epoch": 0.4007919841603168,
+      "grad_norm": 1.3515625,
+      "learning_rate": 2.4727722772277227e-06,
+      "loss": 1.696,
+      "step": 668
+    },
+    {
+      "epoch": 0.4013919721605568,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.47029702970297e-06,
+      "loss": 1.768,
+      "step": 669
+    },
+    {
+      "epoch": 0.4019919601607968,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.4678217821782178e-06,
+      "loss": 1.8123,
+      "step": 670
+    },
+    {
+      "epoch": 0.4025919481610368,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.465346534653465e-06,
+      "loss": 1.8934,
+      "step": 671
+    },
+    {
+      "epoch": 0.4031919361612768,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.462871287128713e-06,
+      "loss": 1.8896,
+      "step": 672
+    },
+    {
+      "epoch": 0.40379192416151677,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.4603960396039602e-06,
+      "loss": 1.9371,
+      "step": 673
+    },
+    {
+      "epoch": 0.40439191216175674,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.457920792079208e-06,
+      "loss": 1.68,
+      "step": 674
+    },
+    {
+      "epoch": 0.40499190016199677,
+      "grad_norm": 1.5,
+      "learning_rate": 2.4554455445544553e-06,
+      "loss": 1.9408,
+      "step": 675
+    },
+    {
+      "epoch": 0.40559188816223674,
+      "grad_norm": 1.4609375,
+      "learning_rate": 2.452970297029703e-06,
+      "loss": 1.7283,
+      "step": 676
+    },
+    {
+      "epoch": 0.40619187616247676,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.4504950495049505e-06,
+      "loss": 1.8666,
+      "step": 677
+    },
+    {
+      "epoch": 0.40679186416271673,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.4480198019801982e-06,
+      "loss": 1.8113,
+      "step": 678
+    },
+    {
+      "epoch": 0.40739185216295676,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.4455445544554456e-06,
+      "loss": 1.8736,
+      "step": 679
+    },
+    {
+      "epoch": 0.40799184016319673,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.4430693069306933e-06,
+      "loss": 1.8091,
+      "step": 680
+    },
+    {
+      "epoch": 0.40859182816343675,
+      "grad_norm": 1.3515625,
+      "learning_rate": 2.4405940594059402e-06,
+      "loss": 1.8982,
+      "step": 681
+    },
+    {
+      "epoch": 0.4091918161636767,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.438118811881188e-06,
+      "loss": 1.8069,
+      "step": 682
+    },
+    {
+      "epoch": 0.4097918041639167,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.4356435643564354e-06,
+      "loss": 1.8585,
+      "step": 683
+    },
+    {
+      "epoch": 0.4103917921641567,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.433168316831683e-06,
+      "loss": 1.9333,
+      "step": 684
+    },
+    {
+      "epoch": 0.4109917801643967,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.4306930693069305e-06,
+      "loss": 1.8872,
+      "step": 685
+    },
+    {
+      "epoch": 0.4115917681646367,
+      "grad_norm": 1.421875,
+      "learning_rate": 2.4282178217821782e-06,
+      "loss": 1.8252,
+      "step": 686
+    },
+    {
+      "epoch": 0.4121917561648767,
+      "grad_norm": 1.3359375,
+      "learning_rate": 2.4257425742574256e-06,
+      "loss": 1.7709,
+      "step": 687
+    },
+    {
+      "epoch": 0.4127917441651167,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.4232673267326733e-06,
+      "loss": 1.8485,
+      "step": 688
+    },
+    {
+      "epoch": 0.4133917321653567,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.4207920792079207e-06,
+      "loss": 1.7962,
+      "step": 689
+    },
+    {
+      "epoch": 0.4139917201655967,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.4183168316831685e-06,
+      "loss": 1.8375,
+      "step": 690
+    },
+    {
+      "epoch": 0.4145917081658367,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.415841584158416e-06,
+      "loss": 1.8245,
+      "step": 691
+    },
+    {
+      "epoch": 0.41519169616607665,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.4133663366336636e-06,
+      "loss": 1.9368,
+      "step": 692
+    },
+    {
+      "epoch": 0.4157916841663167,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.410891089108911e-06,
+      "loss": 1.8899,
+      "step": 693
+    },
+    {
+      "epoch": 0.41639167216655665,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.4084158415841582e-06,
+      "loss": 1.8622,
+      "step": 694
+    },
+    {
+      "epoch": 0.4169916601667967,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.4059405940594056e-06,
+      "loss": 1.77,
+      "step": 695
+    },
+    {
+      "epoch": 0.41759164816703664,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.4034653465346534e-06,
+      "loss": 1.983,
+      "step": 696
+    },
+    {
+      "epoch": 0.41819163616727667,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.4009900990099007e-06,
+      "loss": 1.9289,
+      "step": 697
+    },
+    {
+      "epoch": 0.41879162416751664,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.3985148514851485e-06,
+      "loss": 1.7553,
+      "step": 698
+    },
+    {
+      "epoch": 0.41939161216775667,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.396039603960396e-06,
+      "loss": 1.9037,
+      "step": 699
+    },
+    {
+      "epoch": 0.41999160016799664,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.3935643564356436e-06,
+      "loss": 1.8117,
+      "step": 700
+    },
+    {
+      "epoch": 0.41999160016799664,
+      "eval_loss": 2.0324764251708984,
+      "eval_model_preparation_time": 0.0036,
+      "eval_runtime": 65.1863,
+      "eval_samples_per_second": 153.407,
+      "eval_steps_per_second": 25.573,
+      "step": 700
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1666,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3404255362492662e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}