diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6607 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 939,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003194888178913738,
+      "grad_norm": 16.717142207453772,
+      "learning_rate": 0.0,
+      "loss": 1.805,
+      "step": 1
+    },
+    {
+      "epoch": 0.006389776357827476,
+      "grad_norm": 14.023251259992671,
+      "learning_rate": 1.0638297872340426e-07,
+      "loss": 1.7437,
+      "step": 2
+    },
+    {
+      "epoch": 0.009584664536741214,
+      "grad_norm": 20.233063373741423,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 1.7713,
+      "step": 3
+    },
+    {
+      "epoch": 0.012779552715654952,
+      "grad_norm": 32.40631329434045,
+      "learning_rate": 3.1914893617021275e-07,
+      "loss": 1.8664,
+      "step": 4
+    },
+    {
+      "epoch": 0.01597444089456869,
+      "grad_norm": 11.151298510792111,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 1.7119,
+      "step": 5
+    },
+    {
+      "epoch": 0.019169329073482427,
+      "grad_norm": 16.97709612304111,
+      "learning_rate": 5.319148936170213e-07,
+      "loss": 1.8852,
+      "step": 6
+    },
+    {
+      "epoch": 0.022364217252396165,
+      "grad_norm": 16.479785020928333,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 1.7909,
+      "step": 7
+    },
+    {
+      "epoch": 0.025559105431309903,
+      "grad_norm": 31.984915894604853,
+      "learning_rate": 7.446808510638298e-07,
+      "loss": 1.5985,
+      "step": 8
+    },
+    {
+      "epoch": 0.02875399361022364,
+      "grad_norm": 18.405770787032864,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 1.6649,
+      "step": 9
+    },
+    {
+      "epoch": 0.03194888178913738,
+      "grad_norm": 11.568457035620957,
+      "learning_rate": 9.574468085106384e-07,
+      "loss": 1.6343,
+      "step": 10
+    },
+    {
+      "epoch": 0.03514376996805112,
+      "grad_norm": 11.99369798309996,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 1.7131,
+      "step": 11
+    },
+    {
+      "epoch": 0.038338658146964855,
+      "grad_norm": 16.579277724503285,
+      "learning_rate": 1.170212765957447e-06,
+      "loss": 1.6781,
+      "step": 12
+    },
+    {
+      "epoch": 0.04153354632587859,
+      "grad_norm": 23.280241560149864,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 1.9946,
+      "step": 13
+    },
+    {
+      "epoch": 0.04472843450479233,
+      "grad_norm": 11.069508575430309,
+      "learning_rate": 1.3829787234042555e-06,
+      "loss": 1.7037,
+      "step": 14
+    },
+    {
+      "epoch": 0.04792332268370607,
+      "grad_norm": 11.222580048610657,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 1.5045,
+      "step": 15
+    },
+    {
+      "epoch": 0.051118210862619806,
+      "grad_norm": 9.262248481698279,
+      "learning_rate": 1.595744680851064e-06,
+      "loss": 1.6869,
+      "step": 16
+    },
+    {
+      "epoch": 0.054313099041533544,
+      "grad_norm": 8.164029447660837,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 1.6781,
+      "step": 17
+    },
+    {
+      "epoch": 0.05750798722044728,
+      "grad_norm": 11.070212410234445,
+      "learning_rate": 1.8085106382978727e-06,
+      "loss": 1.5084,
+      "step": 18
+    },
+    {
+      "epoch": 0.06070287539936102,
+      "grad_norm": 18.24579269482639,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 1.6718,
+      "step": 19
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 7.459594379769327,
+      "learning_rate": 2.021276595744681e-06,
+      "loss": 1.4803,
+      "step": 20
+    },
+    {
+      "epoch": 0.0670926517571885,
+      "grad_norm": 10.804667596623585,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 1.451,
+      "step": 21
+    },
+    {
+      "epoch": 0.07028753993610223,
+      "grad_norm": 7.21964134256131,
+      "learning_rate": 2.2340425531914894e-06,
+      "loss": 1.522,
+      "step": 22
+    },
+    {
+      "epoch": 0.07348242811501597,
+      "grad_norm": 10.067355095577458,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 1.4243,
+      "step": 23
+    },
+    {
+      "epoch": 0.07667731629392971,
+      "grad_norm": 16.1259819071595,
+      "learning_rate": 2.446808510638298e-06,
+      "loss": 1.3842,
+      "step": 24
+    },
+    {
+      "epoch": 0.07987220447284345,
+      "grad_norm": 10.371859175411412,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 1.383,
+      "step": 25
+    },
+    {
+      "epoch": 0.08306709265175719,
+      "grad_norm": 8.964914217789572,
+      "learning_rate": 2.6595744680851065e-06,
+      "loss": 1.5774,
+      "step": 26
+    },
+    {
+      "epoch": 0.08626198083067092,
+      "grad_norm": 7.746583868291803,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 1.5022,
+      "step": 27
+    },
+    {
+      "epoch": 0.08945686900958466,
+      "grad_norm": 5.89473714557576,
+      "learning_rate": 2.8723404255319155e-06,
+      "loss": 1.3396,
+      "step": 28
+    },
+    {
+      "epoch": 0.0926517571884984,
+      "grad_norm": 5.02699347301002,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 1.3837,
+      "step": 29
+    },
+    {
+      "epoch": 0.09584664536741214,
+      "grad_norm": 9.371837428436915,
+      "learning_rate": 3.0851063829787237e-06,
+      "loss": 1.5456,
+      "step": 30
+    },
+    {
+      "epoch": 0.09904153354632587,
+      "grad_norm": 11.06547207040692,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 1.4454,
+      "step": 31
+    },
+    {
+      "epoch": 0.10223642172523961,
+      "grad_norm": 9.221564102628951,
+      "learning_rate": 3.297872340425532e-06,
+      "loss": 1.2007,
+      "step": 32
+    },
+    {
+      "epoch": 0.10543130990415335,
+      "grad_norm": 8.929546733679098,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 1.4135,
+      "step": 33
+    },
+    {
+      "epoch": 0.10862619808306709,
+      "grad_norm": 9.211874077709819,
+      "learning_rate": 3.510638297872341e-06,
+      "loss": 1.2891,
+      "step": 34
+    },
+    {
+      "epoch": 0.11182108626198083,
+      "grad_norm": 7.000308016735031,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 1.2162,
+      "step": 35
+    },
+    {
+      "epoch": 0.11501597444089456,
+      "grad_norm": 4.7532035690557475,
+      "learning_rate": 3.723404255319149e-06,
+      "loss": 1.2186,
+      "step": 36
+    },
+    {
+      "epoch": 0.1182108626198083,
+      "grad_norm": 6.27451384888466,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 1.22,
+      "step": 37
+    },
+    {
+      "epoch": 0.12140575079872204,
+      "grad_norm": 5.883795747493129,
+      "learning_rate": 3.936170212765958e-06,
+      "loss": 1.1862,
+      "step": 38
+    },
+    {
+      "epoch": 0.12460063897763578,
+      "grad_norm": 5.585568962155399,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 1.2616,
+      "step": 39
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 6.088301652676503,
+      "learning_rate": 4.148936170212766e-06,
+      "loss": 1.2963,
+      "step": 40
+    },
+    {
+      "epoch": 0.13099041533546327,
+      "grad_norm": 6.236302292423119,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 1.3596,
+      "step": 41
+    },
+    {
+      "epoch": 0.134185303514377,
+      "grad_norm": 7.277813253034166,
+      "learning_rate": 4.361702127659575e-06,
+      "loss": 1.355,
+      "step": 42
+    },
+    {
+      "epoch": 0.13738019169329074,
+      "grad_norm": 6.9466699695085925,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 1.2406,
+      "step": 43
+    },
+    {
+      "epoch": 0.14057507987220447,
+      "grad_norm": 5.400956662876495,
+      "learning_rate": 4.574468085106383e-06,
+      "loss": 1.174,
+      "step": 44
+    },
+    {
+      "epoch": 0.14376996805111822,
+      "grad_norm": 4.824827492352657,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 1.1444,
+      "step": 45
+    },
+    {
+      "epoch": 0.14696485623003194,
+      "grad_norm": 4.13139301261386,
+      "learning_rate": 4.787234042553192e-06,
+      "loss": 0.9866,
+      "step": 46
+    },
+    {
+      "epoch": 0.1501597444089457,
+      "grad_norm": 4.48141844546404,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 1.2958,
+      "step": 47
+    },
+    {
+      "epoch": 0.15335463258785942,
+      "grad_norm": 5.194477479016162,
+      "learning_rate": 5e-06,
+      "loss": 1.279,
+      "step": 48
+    },
+    {
+      "epoch": 0.15654952076677317,
+      "grad_norm": 42.09662909433009,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 1.0737,
+      "step": 49
+    },
+    {
+      "epoch": 0.1597444089456869,
+      "grad_norm": 6.67181650955879,
+      "learning_rate": 5.212765957446809e-06,
+      "loss": 1.174,
+      "step": 50
+    },
+    {
+      "epoch": 0.16293929712460065,
+      "grad_norm": 4.305530439030815,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 1.1243,
+      "step": 51
+    },
+    {
+      "epoch": 0.16613418530351437,
+      "grad_norm": 5.8404661477440545,
+      "learning_rate": 5.425531914893617e-06,
+      "loss": 1.2639,
+      "step": 52
+    },
+    {
+      "epoch": 0.16932907348242812,
+      "grad_norm": 5.257506385774717,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 1.0899,
+      "step": 53
+    },
+    {
+      "epoch": 0.17252396166134185,
+      "grad_norm": 4.775620240966863,
+      "learning_rate": 5.638297872340426e-06,
+      "loss": 1.0457,
+      "step": 54
+    },
+    {
+      "epoch": 0.1757188498402556,
+      "grad_norm": 5.764147717959987,
+      "learning_rate": 5.744680851063831e-06,
+      "loss": 1.1835,
+      "step": 55
+    },
+    {
+      "epoch": 0.17891373801916932,
+      "grad_norm": 4.098061876130292,
+      "learning_rate": 5.851063829787235e-06,
+      "loss": 0.9886,
+      "step": 56
+    },
+    {
+      "epoch": 0.18210862619808307,
+      "grad_norm": 4.179637345679237,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 1.3009,
+      "step": 57
+    },
+    {
+      "epoch": 0.1853035143769968,
+      "grad_norm": 3.9920034272044003,
+      "learning_rate": 6.063829787234044e-06,
+      "loss": 1.1074,
+      "step": 58
+    },
+    {
+      "epoch": 0.18849840255591055,
+      "grad_norm": 4.408578981594532,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 1.1697,
+      "step": 59
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 4.325540588859466,
+      "learning_rate": 6.276595744680851e-06,
+      "loss": 1.0138,
+      "step": 60
+    },
+    {
+      "epoch": 0.19488817891373802,
+      "grad_norm": 5.837072539573206,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 1.1069,
+      "step": 61
+    },
+    {
+      "epoch": 0.19808306709265175,
+      "grad_norm": 3.4143305874240513,
+      "learning_rate": 6.48936170212766e-06,
+      "loss": 1.0702,
+      "step": 62
+    },
+    {
+      "epoch": 0.2012779552715655,
+      "grad_norm": 5.159078364222101,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 1.2549,
+      "step": 63
+    },
+    {
+      "epoch": 0.20447284345047922,
+      "grad_norm": 3.6348541737787547,
+      "learning_rate": 6.702127659574469e-06,
+      "loss": 1.0914,
+      "step": 64
+    },
+    {
+      "epoch": 0.20766773162939298,
+      "grad_norm": 3.6398934157968013,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 1.0742,
+      "step": 65
+    },
+    {
+      "epoch": 0.2108626198083067,
+      "grad_norm": 3.967296670527451,
+      "learning_rate": 6.914893617021278e-06,
+      "loss": 1.2861,
+      "step": 66
+    },
+    {
+      "epoch": 0.21405750798722045,
+      "grad_norm": 6.021890770908592,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 1.0945,
+      "step": 67
+    },
+    {
+      "epoch": 0.21725239616613418,
+      "grad_norm": 3.290462621217852,
+      "learning_rate": 7.127659574468085e-06,
+      "loss": 1.1113,
+      "step": 68
+    },
+    {
+      "epoch": 0.22044728434504793,
+      "grad_norm": 4.588489755387563,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 1.0304,
+      "step": 69
+    },
+    {
+      "epoch": 0.22364217252396165,
+      "grad_norm": 4.035382414781595,
+      "learning_rate": 7.340425531914894e-06,
+      "loss": 1.1864,
+      "step": 70
+    },
+    {
+      "epoch": 0.2268370607028754,
+      "grad_norm": 4.04715380563845,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 1.2254,
+      "step": 71
+    },
+    {
+      "epoch": 0.23003194888178913,
+      "grad_norm": 3.829994440351725,
+      "learning_rate": 7.553191489361703e-06,
+      "loss": 1.1626,
+      "step": 72
+    },
+    {
+      "epoch": 0.23322683706070288,
+      "grad_norm": 4.479079459670903,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 1.1799,
+      "step": 73
+    },
+    {
+      "epoch": 0.2364217252396166,
+      "grad_norm": 3.678691371428723,
+      "learning_rate": 7.765957446808511e-06,
+      "loss": 1.2507,
+      "step": 74
+    },
+    {
+      "epoch": 0.23961661341853036,
+      "grad_norm": 4.437784808726001,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 1.2643,
+      "step": 75
+    },
+    {
+      "epoch": 0.24281150159744408,
+      "grad_norm": 3.9846462409584418,
+      "learning_rate": 7.97872340425532e-06,
+      "loss": 1.1032,
+      "step": 76
+    },
+    {
+      "epoch": 0.24600638977635783,
+      "grad_norm": 6.940960461152686,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 0.9179,
+      "step": 77
+    },
+    {
+      "epoch": 0.24920127795527156,
+      "grad_norm": 4.481026432045363,
+      "learning_rate": 8.191489361702128e-06,
+      "loss": 1.2025,
+      "step": 78
+    },
+    {
+      "epoch": 0.2523961661341853,
+      "grad_norm": 3.8370366107478984,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 1.1794,
+      "step": 79
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 4.025725358994797,
+      "learning_rate": 8.404255319148937e-06,
+      "loss": 1.232,
+      "step": 80
+    },
+    {
+      "epoch": 0.25878594249201275,
+      "grad_norm": 4.386666891431395,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 1.1926,
+      "step": 81
+    },
+    {
+      "epoch": 0.26198083067092653,
+      "grad_norm": 4.2058282949121475,
+      "learning_rate": 8.617021276595746e-06,
+      "loss": 1.1901,
+      "step": 82
+    },
+    {
+      "epoch": 0.26517571884984026,
+      "grad_norm": 4.120218722147704,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 1.2094,
+      "step": 83
+    },
+    {
+      "epoch": 0.268370607028754,
+      "grad_norm": 4.325967470086441,
+      "learning_rate": 8.829787234042555e-06,
+      "loss": 1.2703,
+      "step": 84
+    },
+    {
+      "epoch": 0.2715654952076677,
+      "grad_norm": 3.9299696349676307,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 1.1243,
+      "step": 85
+    },
+    {
+      "epoch": 0.2747603833865815,
+      "grad_norm": 3.9542166301664414,
+      "learning_rate": 9.042553191489362e-06,
+      "loss": 1.0538,
+      "step": 86
+    },
+    {
+      "epoch": 0.2779552715654952,
+      "grad_norm": 4.358143235123361,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 1.1565,
+      "step": 87
+    },
+    {
+      "epoch": 0.28115015974440893,
+      "grad_norm": 3.55899466233284,
+      "learning_rate": 9.255319148936171e-06,
+      "loss": 1.08,
+      "step": 88
+    },
+    {
+      "epoch": 0.28434504792332266,
+      "grad_norm": 4.223895882766692,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 1.0292,
+      "step": 89
+    },
+    {
+      "epoch": 0.28753993610223644,
+      "grad_norm": 3.759821046515852,
+      "learning_rate": 9.46808510638298e-06,
+      "loss": 1.0494,
+      "step": 90
+    },
+    {
+      "epoch": 0.29073482428115016,
+      "grad_norm": 3.851448850237355,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 1.2499,
+      "step": 91
+    },
+    {
+      "epoch": 0.2939297124600639,
+      "grad_norm": 3.747936281280683,
+      "learning_rate": 9.680851063829787e-06,
+      "loss": 1.0605,
+      "step": 92
+    },
+    {
+      "epoch": 0.2971246006389776,
+      "grad_norm": 4.596810997980656,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 0.9667,
+      "step": 93
+    },
+    {
+      "epoch": 0.3003194888178914,
+      "grad_norm": 3.9850160361720692,
+      "learning_rate": 9.893617021276596e-06,
+      "loss": 1.135,
+      "step": 94
+    },
+    {
+      "epoch": 0.3035143769968051,
+      "grad_norm": 3.8349397037787196,
+      "learning_rate": 1e-05,
+      "loss": 1.1518,
+      "step": 95
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 4.883098045146262,
+      "learning_rate": 9.999965443811378e-06,
+      "loss": 1.0875,
+      "step": 96
+    },
+    {
+      "epoch": 0.30990415335463256,
+      "grad_norm": 3.623618945611754,
+      "learning_rate": 9.999861775723162e-06,
+      "loss": 1.04,
+      "step": 97
+    },
+    {
+      "epoch": 0.31309904153354634,
+      "grad_norm": 3.8417936721502883,
+      "learning_rate": 9.999688997168301e-06,
+      "loss": 0.7891,
+      "step": 98
+    },
+    {
+      "epoch": 0.31629392971246006,
+      "grad_norm": 4.401091520825806,
+      "learning_rate": 9.999447110535026e-06,
+      "loss": 1.0505,
+      "step": 99
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 13.029159973057038,
+      "learning_rate": 9.999136119166803e-06,
+      "loss": 1.1401,
+      "step": 100
+    },
+    {
+      "epoch": 0.3226837060702875,
+      "grad_norm": 4.423887269496716,
+      "learning_rate": 9.998756027362308e-06,
+      "loss": 1.0174,
+      "step": 101
+    },
+    {
+      "epoch": 0.3258785942492013,
+      "grad_norm": 3.7980999538112408,
+      "learning_rate": 9.99830684037535e-06,
+      "loss": 1.1718,
+      "step": 102
+    },
+    {
+      "epoch": 0.329073482428115,
+      "grad_norm": 3.6898294868985877,
+      "learning_rate": 9.9977885644148e-06,
+      "loss": 1.1285,
+      "step": 103
+    },
+    {
+      "epoch": 0.33226837060702874,
+      "grad_norm": 3.9597869172310833,
+      "learning_rate": 9.997201206644522e-06,
+      "loss": 1.1038,
+      "step": 104
+    },
+    {
+      "epoch": 0.3354632587859425,
+      "grad_norm": 4.902853308860501,
+      "learning_rate": 9.99654477518325e-06,
+      "loss": 1.1963,
+      "step": 105
+    },
+    {
+      "epoch": 0.33865814696485624,
+      "grad_norm": 4.044520667607736,
+      "learning_rate": 9.995819279104494e-06,
+      "loss": 1.1649,
+      "step": 106
+    },
+    {
+      "epoch": 0.34185303514376997,
+      "grad_norm": 4.708076049334671,
+      "learning_rate": 9.995024728436402e-06,
+      "loss": 1.335,
+      "step": 107
+    },
+    {
+      "epoch": 0.3450479233226837,
+      "grad_norm": 4.312023697476172,
+      "learning_rate": 9.994161134161635e-06,
+      "loss": 1.0715,
+      "step": 108
+    },
+    {
+      "epoch": 0.34824281150159747,
+      "grad_norm": 4.541045752400942,
+      "learning_rate": 9.993228508217201e-06,
+      "loss": 1.2724,
+      "step": 109
+    },
+    {
+      "epoch": 0.3514376996805112,
+      "grad_norm": 3.966018141765597,
+      "learning_rate": 9.9922268634943e-06,
+      "loss": 0.9365,
+      "step": 110
+    },
+    {
+      "epoch": 0.3546325878594249,
+      "grad_norm": 4.335596202564225,
+      "learning_rate": 9.991156213838143e-06,
+      "loss": 0.9651,
+      "step": 111
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 4.155482161962993,
+      "learning_rate": 9.990016574047757e-06,
+      "loss": 1.2132,
+      "step": 112
+    },
+    {
+      "epoch": 0.3610223642172524,
+      "grad_norm": 3.8733639154188433,
+      "learning_rate": 9.988807959875785e-06,
+      "loss": 1.0806,
+      "step": 113
+    },
+    {
+      "epoch": 0.36421725239616615,
+      "grad_norm": 3.3146945824625393,
+      "learning_rate": 9.987530388028269e-06,
+      "loss": 1.2565,
+      "step": 114
+    },
+    {
+      "epoch": 0.36741214057507987,
+      "grad_norm": 6.051307642992494,
+      "learning_rate": 9.986183876164412e-06,
+      "loss": 1.124,
+      "step": 115
+    },
+    {
+      "epoch": 0.3706070287539936,
+      "grad_norm": 3.4794884296886806,
+      "learning_rate": 9.984768442896342e-06,
+      "loss": 1.1516,
+      "step": 116
+    },
+    {
+      "epoch": 0.3738019169329074,
+      "grad_norm": 4.196373384697492,
+      "learning_rate": 9.983284107788852e-06,
+      "loss": 1.1689,
+      "step": 117
+    },
+    {
+      "epoch": 0.3769968051118211,
+      "grad_norm": 4.552086648154905,
+      "learning_rate": 9.981730891359123e-06,
+      "loss": 1.1735,
+      "step": 118
+    },
+    {
+      "epoch": 0.3801916932907348,
+      "grad_norm": 3.913371024198064,
+      "learning_rate": 9.980108815076456e-06,
+      "loss": 1.1067,
+      "step": 119
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 3.912114563097618,
+      "learning_rate": 9.978417901361958e-06,
+      "loss": 1.1209,
+      "step": 120
+    },
+    {
+      "epoch": 0.3865814696485623,
+      "grad_norm": 4.7332806426376095,
+      "learning_rate": 9.976658173588244e-06,
+      "loss": 1.3591,
+      "step": 121
+    },
+    {
+      "epoch": 0.38977635782747605,
+      "grad_norm": 3.9207746825119845,
+      "learning_rate": 9.974829656079106e-06,
+      "loss": 1.2432,
+      "step": 122
+    },
+    {
+      "epoch": 0.3929712460063898,
+      "grad_norm": 4.132835210944295,
+      "learning_rate": 9.972932374109184e-06,
+      "loss": 1.2292,
+      "step": 123
+    },
+    {
+      "epoch": 0.3961661341853035,
+      "grad_norm": 3.894731189939478,
+      "learning_rate": 9.97096635390361e-06,
+      "loss": 1.2887,
+      "step": 124
+    },
+    {
+      "epoch": 0.3993610223642173,
+      "grad_norm": 4.355097360308356,
+      "learning_rate": 9.968931622637652e-06,
+      "loss": 1.1217,
+      "step": 125
+    },
+    {
+      "epoch": 0.402555910543131,
+      "grad_norm": 4.139296952481508,
+      "learning_rate": 9.966828208436332e-06,
+      "loss": 1.2401,
+      "step": 126
+    },
+    {
+      "epoch": 0.4057507987220447,
+      "grad_norm": 3.818968828059459,
+      "learning_rate": 9.96465614037404e-06,
+      "loss": 1.0963,
+      "step": 127
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 3.7926778781572894,
+      "learning_rate": 9.962415448474134e-06,
+      "loss": 1.2293,
+      "step": 128
+    },
+    {
+      "epoch": 0.41214057507987223,
+      "grad_norm": 4.1360290936180535,
+      "learning_rate": 9.960106163708522e-06,
+      "loss": 1.2511,
+      "step": 129
+    },
+    {
+      "epoch": 0.41533546325878595,
+      "grad_norm": 3.8907965069506654,
+      "learning_rate": 9.95772831799724e-06,
+      "loss": 1.153,
+      "step": 130
+    },
+    {
+      "epoch": 0.4185303514376997,
+      "grad_norm": 4.077376785667901,
+      "learning_rate": 9.955281944207998e-06,
+      "loss": 1.0657,
+      "step": 131
+    },
+    {
+      "epoch": 0.4217252396166134,
+      "grad_norm": 4.148299433167882,
+      "learning_rate": 9.95276707615574e-06,
+      "loss": 1.2122,
+      "step": 132
+    },
+    {
+      "epoch": 0.4249201277955272,
+      "grad_norm": 3.6998606278397728,
+      "learning_rate": 9.950183748602164e-06,
+      "loss": 1.1245,
+      "step": 133
+    },
+    {
+      "epoch": 0.4281150159744409,
+      "grad_norm": 3.6878076710266012,
+      "learning_rate": 9.947531997255256e-06,
+      "loss": 1.321,
+      "step": 134
+    },
+    {
+      "epoch": 0.43130990415335463,
+      "grad_norm": 4.343087297558039,
+      "learning_rate": 9.944811858768782e-06,
+      "loss": 1.1943,
+      "step": 135
+    },
+    {
+      "epoch": 0.43450479233226835,
+      "grad_norm": 3.50047150176892,
+      "learning_rate": 9.94202337074179e-06,
+      "loss": 1.0774,
+      "step": 136
+    },
+    {
+      "epoch": 0.43769968051118213,
+      "grad_norm": 4.616763355783899,
+      "learning_rate": 9.939166571718086e-06,
+      "loss": 1.214,
+      "step": 137
+    },
+    {
+      "epoch": 0.44089456869009586,
+      "grad_norm": 4.53205696332133,
+      "learning_rate": 9.936241501185706e-06,
+      "loss": 1.1955,
+      "step": 138
+    },
+    {
+      "epoch": 0.4440894568690096,
+      "grad_norm": 4.411589524995418,
+      "learning_rate": 9.933248199576366e-06,
+      "loss": 1.3724,
+      "step": 139
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 4.477761797640728,
+      "learning_rate": 9.930186708264902e-06,
+      "loss": 1.0907,
+      "step": 140
+    },
+    {
+      "epoch": 0.4504792332268371,
+      "grad_norm": 3.9586204953226605,
+      "learning_rate": 9.927057069568704e-06,
+      "loss": 1.1487,
+      "step": 141
+    },
+    {
+      "epoch": 0.4536741214057508,
+      "grad_norm": 3.56582897022002,
+      "learning_rate": 9.923859326747125e-06,
+      "loss": 1.1551,
+      "step": 142
+    },
+    {
+      "epoch": 0.45686900958466453,
+      "grad_norm": 4.06029005798942,
+      "learning_rate": 9.920593524000887e-06,
+      "loss": 1.2013,
+      "step": 143
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 4.064412435822573,
+      "learning_rate": 9.917259706471469e-06,
+      "loss": 1.0198,
+      "step": 144
+    },
+    {
+      "epoch": 0.46325878594249204,
+      "grad_norm": 3.37085812413093,
+      "learning_rate": 9.913857920240481e-06,
+      "loss": 1.1357,
+      "step": 145
+    },
+    {
+      "epoch": 0.46645367412140576,
+      "grad_norm": 3.404689910480565,
+      "learning_rate": 9.91038821232903e-06,
+      "loss": 1.1617,
+      "step": 146
+    },
+    {
+      "epoch": 0.4696485623003195,
+      "grad_norm": 3.5984111613146545,
+      "learning_rate": 9.906850630697068e-06,
+      "loss": 1.2702,
+      "step": 147
+    },
+    {
+      "epoch": 0.4728434504792332,
+      "grad_norm": 4.020912214133055,
+      "learning_rate": 9.903245224242732e-06,
+      "loss": 1.1822,
+      "step": 148
+    },
+    {
+      "epoch": 0.476038338658147,
+      "grad_norm": 3.803789421529034,
+      "learning_rate": 9.899572042801662e-06,
+      "loss": 1.0678,
+      "step": 149
+    },
+    {
+      "epoch": 0.4792332268370607,
+      "grad_norm": 3.7451070640757163,
+      "learning_rate": 9.895831137146319e-06,
+      "loss": 1.1907,
+      "step": 150
+    },
+    {
+      "epoch": 0.48242811501597443,
+      "grad_norm": 3.505027126947319,
+      "learning_rate": 9.89202255898528e-06,
+      "loss": 1.1625,
+      "step": 151
+    },
+    {
+      "epoch": 0.48562300319488816,
+      "grad_norm": 3.8633072484745927,
+      "learning_rate": 9.888146360962523e-06,
+      "loss": 1.2423,
+      "step": 152
+    },
+    {
+      "epoch": 0.48881789137380194,
+      "grad_norm": 4.4981059716343195,
+      "learning_rate": 9.8842025966567e-06,
+      "loss": 1.2166,
+      "step": 153
+    },
+    {
+      "epoch": 0.49201277955271566,
+      "grad_norm": 3.6721067111812715,
+      "learning_rate": 9.880191320580396e-06,
+      "loss": 1.0802,
+      "step": 154
+    },
+    {
+      "epoch": 0.4952076677316294,
+      "grad_norm": 3.636110274717446,
+      "learning_rate": 9.876112588179378e-06,
+      "loss": 1.0047,
+      "step": 155
+    },
+    {
+      "epoch": 0.4984025559105431,
+      "grad_norm": 5.322021427043842,
+      "learning_rate": 9.87196645583182e-06,
+      "loss": 1.0806,
+      "step": 156
+    },
+    {
+      "epoch": 0.5015974440894568,
+      "grad_norm": 3.611129253446001,
+      "learning_rate": 9.86775298084754e-06,
+      "loss": 1.1326,
+      "step": 157
+    },
+    {
+      "epoch": 0.5047923322683706,
+      "grad_norm": 3.6538957956039533,
+      "learning_rate": 9.863472221467189e-06,
+      "loss": 1.1356,
+      "step": 158
+    },
+    {
+      "epoch": 0.5079872204472844,
+      "grad_norm": 3.849459831501915,
+      "learning_rate": 9.85912423686146e-06,
+      "loss": 0.9944,
+      "step": 159
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 3.5672182387089078,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 1.1254,
+      "step": 160
+    },
+    {
+      "epoch": 0.5143769968051118,
+      "grad_norm": 3.7738669234991145,
+      "learning_rate": 9.850226833301893e-06,
+      "loss": 1.1222,
+      "step": 161
+    },
+    {
+      "epoch": 0.5175718849840255,
+      "grad_norm": 3.854093721258117,
+      "learning_rate": 9.8456775373322e-06,
+      "loss": 1.3054,
+      "step": 162
+    },
+    {
+      "epoch": 0.5207667731629393,
+      "grad_norm": 3.3696596715795595,
+      "learning_rate": 9.841061262103713e-06,
+      "loss": 1.1182,
+      "step": 163
+    },
+    {
+      "epoch": 0.5239616613418531,
+      "grad_norm": 4.3018619746731455,
+      "learning_rate": 9.836378071424782e-06,
+      "loss": 1.1178,
+      "step": 164
+    },
+    {
+      "epoch": 0.5271565495207667,
+      "grad_norm": 3.757001775301693,
+      "learning_rate": 9.831628030028698e-06,
+      "loss": 0.9919,
+      "step": 165
+    },
+    {
+      "epoch": 0.5303514376996805,
+      "grad_norm": 3.447400890761457,
+      "learning_rate": 9.826811203572785e-06,
+      "loss": 1.0795,
+      "step": 166
+    },
+    {
+      "epoch": 0.5335463258785943,
+      "grad_norm": 3.524139827921683,
+      "learning_rate": 9.821927658637518e-06,
+      "loss": 1.0783,
+      "step": 167
+    },
+    {
+      "epoch": 0.536741214057508,
+      "grad_norm": 3.5324904560401187,
+      "learning_rate": 9.81697746272557e-06,
+      "loss": 1.1246,
+      "step": 168
+    },
+    {
+      "epoch": 0.5399361022364217,
+      "grad_norm": 5.07674793145727,
+      "learning_rate": 9.811960684260907e-06,
+      "loss": 1.2482,
+      "step": 169
+    },
+    {
+      "epoch": 0.5431309904153354,
+      "grad_norm": 3.9093911845048175,
+      "learning_rate": 9.80687739258782e-06,
+      "loss": 1.3509,
+      "step": 170
+    },
+    {
+      "epoch": 0.5463258785942492,
+      "grad_norm": 3.8993712172541235,
+      "learning_rate": 9.801727657969988e-06,
+      "loss": 1.1233,
+      "step": 171
+    },
+    {
+      "epoch": 0.549520766773163,
+      "grad_norm": 3.4393141842942287,
+      "learning_rate": 9.796511551589492e-06,
+      "loss": 1.1965,
+      "step": 172
+    },
+    {
+      "epoch": 0.5527156549520766,
+      "grad_norm": 3.679560530304304,
+      "learning_rate": 9.791229145545832e-06,
+      "loss": 1.0492,
+      "step": 173
+    },
+    {
+      "epoch": 0.5559105431309904,
+      "grad_norm": 3.1309410887255478,
+      "learning_rate": 9.785880512854937e-06,
+      "loss": 1.0509,
+      "step": 174
+    },
+    {
+      "epoch": 0.5591054313099042,
+      "grad_norm": 4.218070130605334,
+      "learning_rate": 9.78046572744815e-06,
+      "loss": 1.1435,
+      "step": 175
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 4.077312544218879,
+      "learning_rate": 9.77498486417121e-06,
+      "loss": 1.0912,
+      "step": 176
+    },
+    {
+      "epoch": 0.5654952076677316,
+      "grad_norm": 3.8127715243949067,
+      "learning_rate": 9.769437998783216e-06,
+      "loss": 1.2658,
+      "step": 177
+    },
+    {
+      "epoch": 0.5686900958466453,
+      "grad_norm": 4.327358073190962,
+      "learning_rate": 9.763825207955577e-06,
+      "loss": 1.2511,
+      "step": 178
+    },
+    {
+      "epoch": 0.5718849840255591,
+      "grad_norm": 3.108421475702593,
+      "learning_rate": 9.758146569270957e-06,
+      "loss": 1.158,
+      "step": 179
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 3.0665371887942396,
+      "learning_rate": 9.7524021612222e-06,
+      "loss": 1.0504,
+      "step": 180
+    },
+    {
+      "epoch": 0.5782747603833865,
+      "grad_norm": 4.452192164910537,
+      "learning_rate": 9.746592063211247e-06,
+      "loss": 0.9544,
+      "step": 181
+    },
+    {
+      "epoch": 0.5814696485623003,
+      "grad_norm": 4.0461041827089534,
+      "learning_rate": 9.74071635554803e-06,
+      "loss": 1.1458,
+      "step": 182
+    },
+    {
+      "epoch": 0.5846645367412141,
+      "grad_norm": 3.7945452752886584,
+      "learning_rate": 9.73477511944938e-06,
+      "loss": 1.0422,
+      "step": 183
+    },
+    {
+      "epoch": 0.5878594249201278,
+      "grad_norm": 3.5218405647695885,
+      "learning_rate": 9.728768437037882e-06,
+      "loss": 1.3021,
+      "step": 184
+    },
+    {
+      "epoch": 0.5910543130990416,
+      "grad_norm": 3.7387399313812177,
+      "learning_rate": 9.722696391340762e-06,
+      "loss": 1.1793,
+      "step": 185
+    },
+    {
+      "epoch": 0.5942492012779552,
+      "grad_norm": 3.668335801414106,
+      "learning_rate": 9.716559066288716e-06,
+      "loss": 1.0721,
+      "step": 186
+    },
+    {
+      "epoch": 0.597444089456869,
+      "grad_norm": 3.9375276818066904,
+      "learning_rate": 9.710356546714774e-06,
+      "loss": 1.2097,
+      "step": 187
+    },
+    {
+      "epoch": 0.6006389776357828,
+      "grad_norm": 3.442142918013285,
+      "learning_rate": 9.704088918353108e-06,
+      "loss": 0.9673,
+      "step": 188
+    },
+    {
+      "epoch": 0.6038338658146964,
+      "grad_norm": 3.4527926423453805,
+      "learning_rate": 9.697756267837856e-06,
+      "loss": 1.2348,
+      "step": 189
+    },
+    {
+      "epoch": 0.6070287539936102,
+      "grad_norm": 4.458800799597442,
+      "learning_rate": 9.691358682701927e-06,
+      "loss": 1.0925,
+      "step": 190
+    },
+    {
+      "epoch": 0.610223642172524,
+      "grad_norm": 3.923108354585714,
+      "learning_rate": 9.684896251375784e-06,
+      "loss": 1.2055,
+      "step": 191
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 3.7163689003708034,
+      "learning_rate": 9.678369063186224e-06,
+      "loss": 1.1835,
+      "step": 192
+    },
+    {
+      "epoch": 0.6166134185303515,
+      "grad_norm": 4.249443443749534,
+      "learning_rate": 9.671777208355146e-06,
+      "loss": 1.1022,
+      "step": 193
+    },
+    {
+      "epoch": 0.6198083067092651,
+      "grad_norm": 3.7680695876700816,
+      "learning_rate": 9.665120777998303e-06,
+      "loss": 1.1215,
+      "step": 194
+    },
+    {
+      "epoch": 0.6230031948881789,
+      "grad_norm": 3.8560350741289517,
+      "learning_rate": 9.658399864124037e-06,
+      "loss": 1.4127,
+      "step": 195
+    },
+    {
+      "epoch": 0.6261980830670927,
+      "grad_norm": 3.4026835096292385,
+      "learning_rate": 9.65161455963202e-06,
+      "loss": 1.2741,
+      "step": 196
+    },
+    {
+      "epoch": 0.6293929712460063,
+      "grad_norm": 3.773795606779852,
+      "learning_rate": 9.64476495831195e-06,
+      "loss": 1.3382,
+      "step": 197
+    },
+    {
+      "epoch": 0.6325878594249201,
+      "grad_norm": 3.2984077688734255,
+      "learning_rate": 9.637851154842279e-06,
+      "loss": 1.0254,
+      "step": 198
+    },
+    {
+      "epoch": 0.6357827476038339,
+      "grad_norm": 4.646332401988159,
+      "learning_rate": 9.630873244788884e-06,
+      "loss": 1.1941,
+      "step": 199
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 4.060068044776168,
+      "learning_rate": 9.623831324603755e-06,
+      "loss": 1.1995,
+      "step": 200
+    },
+    {
+      "epoch": 0.6421725239616614,
+      "grad_norm": 3.5889106845325074,
+      "learning_rate": 9.61672549162366e-06,
+      "loss": 0.9732,
+      "step": 201
+    },
+    {
+      "epoch": 0.645367412140575,
+      "grad_norm": 3.036851585403188,
+      "learning_rate": 9.6095558440688e-06,
+      "loss": 1.0368,
+      "step": 202
+    },
+    {
+      "epoch": 0.6485623003194888,
+      "grad_norm": 4.160366314906386,
+      "learning_rate": 9.602322481041457e-06,
+      "loss": 1.1353,
+      "step": 203
+    },
+    {
+      "epoch": 0.6517571884984026,
+      "grad_norm": 3.660461317633582,
+      "learning_rate": 9.595025502524609e-06,
+      "loss": 1.1169,
+      "step": 204
+    },
+    {
+      "epoch": 0.6549520766773163,
+      "grad_norm": 4.880222525588506,
+      "learning_rate": 9.587665009380565e-06,
+      "loss": 1.1977,
+      "step": 205
+    },
+    {
+      "epoch": 0.65814696485623,
+      "grad_norm": 4.28269717578457,
+      "learning_rate": 9.580241103349562e-06,
+      "loss": 1.1611,
+      "step": 206
+    },
+    {
+      "epoch": 0.6613418530351438,
+      "grad_norm": 3.359632639410027,
+      "learning_rate": 9.572753887048353e-06,
+      "loss": 1.3119,
+      "step": 207
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 4.1882736364682325,
+      "learning_rate": 9.565203463968808e-06,
+      "loss": 1.2341,
+      "step": 208
+    },
+    {
+      "epoch": 0.6677316293929713,
+      "grad_norm": 4.251723023763134,
+      "learning_rate": 9.557589938476462e-06,
+      "loss": 1.181,
+      "step": 209
+    },
+    {
+      "epoch": 0.670926517571885,
+      "grad_norm": 3.6989491360546873,
+      "learning_rate": 9.549913415809084e-06,
+      "loss": 1.1145,
+      "step": 210
+    },
+    {
+      "epoch": 0.6741214057507987,
+      "grad_norm": 3.720983119620804,
+      "learning_rate": 9.542174002075221e-06,
+      "loss": 1.1195,
+      "step": 211
+    },
+    {
+      "epoch": 0.6773162939297125,
+      "grad_norm": 3.544471454032533,
+      "learning_rate": 9.534371804252727e-06,
+      "loss": 0.9599,
+      "step": 212
+    },
+    {
+      "epoch": 0.6805111821086262,
+      "grad_norm": 3.372468663897121,
+      "learning_rate": 9.526506930187294e-06,
+      "loss": 1.0417,
+      "step": 213
+    },
+    {
+      "epoch": 0.6837060702875399,
+      "grad_norm": 3.840460242760645,
+      "learning_rate": 9.518579488590947e-06,
+      "loss": 1.0221,
+      "step": 214
+    },
+    {
+      "epoch": 0.6869009584664537,
+      "grad_norm": 3.4635380391049937,
+      "learning_rate": 9.510589589040554e-06,
+      "loss": 1.0547,
+      "step": 215
+    },
+    {
+      "epoch": 0.6900958466453674,
+      "grad_norm": 3.754811039613256,
+      "learning_rate": 9.502537341976305e-06,
+      "loss": 1.2437,
+      "step": 216
+    },
+    {
+      "epoch": 0.6932907348242812,
+      "grad_norm": 4.045939263058207,
+      "learning_rate": 9.494422858700188e-06,
+      "loss": 1.1614,
+      "step": 217
+    },
+    {
+      "epoch": 0.6964856230031949,
+      "grad_norm": 3.9450417222440977,
+      "learning_rate": 9.48624625137445e-06,
+      "loss": 1.1075,
+      "step": 218
+    },
+    {
+      "epoch": 0.6996805111821086,
+      "grad_norm": 4.413847851519682,
+      "learning_rate": 9.478007633020043e-06,
+      "loss": 1.1221,
+      "step": 219
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 3.388141122832844,
+      "learning_rate": 9.469707117515068e-06,
+      "loss": 1.0616,
+      "step": 220
+    },
+    {
+      "epoch": 0.7060702875399361,
+      "grad_norm": 3.746490582181775,
+      "learning_rate": 9.461344819593194e-06,
+      "loss": 1.1685,
+      "step": 221
+    },
+    {
+      "epoch": 0.7092651757188498,
+      "grad_norm": 3.414204371068498,
+      "learning_rate": 9.452920854842085e-06,
+      "loss": 1.2356,
+      "step": 222
+    },
+    {
+      "epoch": 0.7124600638977636,
+      "grad_norm": 3.906351625603022,
+      "learning_rate": 9.44443533970178e-06,
+      "loss": 1.1701,
+      "step": 223
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 4.149620037916644,
+      "learning_rate": 9.435888391463108e-06,
+      "loss": 1.0644,
+      "step": 224
+    },
+    {
+      "epoch": 0.7188498402555911,
+      "grad_norm": 4.071378722274369,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 1.1801,
+      "step": 225
+    },
+    {
+      "epoch": 0.7220447284345048,
+      "grad_norm": 4.080246539187525,
+      "learning_rate": 9.418610669098114e-06,
+      "loss": 1.0989,
+      "step": 226
+    },
+    {
+      "epoch": 0.7252396166134185,
+      "grad_norm": 3.5277259743641127,
+      "learning_rate": 9.409880133792684e-06,
+      "loss": 1.2842,
+      "step": 227
+    },
+    {
+      "epoch": 0.7284345047923323,
+      "grad_norm": 4.997225962376027,
+      "learning_rate": 9.40108864302737e-06,
+      "loss": 1.2365,
+      "step": 228
+    },
+    {
+      "epoch": 0.731629392971246,
+      "grad_norm": 3.7691450950069534,
+      "learning_rate": 9.392236318322339e-06,
+      "loss": 0.9714,
+      "step": 229
+    },
+    {
+      "epoch": 0.7348242811501597,
+      "grad_norm": 4.4950654636833915,
+      "learning_rate": 9.383323282038632e-06,
+      "loss": 1.2284,
+      "step": 230
+    },
+    {
+      "epoch": 0.7380191693290735,
+      "grad_norm": 4.0050206239005455,
+      "learning_rate": 9.374349657376473e-06,
+      "loss": 1.2436,
+      "step": 231
+    },
+    {
+      "epoch": 0.7412140575079872,
+      "grad_norm": 4.3522829687961035,
+      "learning_rate": 9.365315568373569e-06,
+      "loss": 1.2282,
+      "step": 232
+    },
+    {
+      "epoch": 0.744408945686901,
+      "grad_norm": 3.542512103159345,
+      "learning_rate": 9.356221139903395e-06,
+      "loss": 1.0976,
+      "step": 233
+    },
+    {
+      "epoch": 0.7476038338658147,
+      "grad_norm": 3.592602507675251,
+      "learning_rate": 9.347066497673462e-06,
+      "loss": 1.2434,
+      "step": 234
+    },
+    {
+      "epoch": 0.7507987220447284,
+      "grad_norm": 3.64500718148932,
+      "learning_rate": 9.337851768223589e-06,
+      "loss": 1.0611,
+      "step": 235
+    },
+    {
+      "epoch": 0.7539936102236422,
+      "grad_norm": 3.6722983671352236,
+      "learning_rate": 9.328577078924151e-06,
+      "loss": 1.1983,
+      "step": 236
+    },
+    {
+      "epoch": 0.7571884984025559,
+      "grad_norm": 4.346034040255591,
+      "learning_rate": 9.319242557974306e-06,
+      "loss": 1.0993,
+      "step": 237
+    },
+    {
+      "epoch": 0.7603833865814696,
+      "grad_norm": 3.379176928406739,
+      "learning_rate": 9.309848334400247e-06,
+      "loss": 1.1768,
+      "step": 238
+    },
+    {
+      "epoch": 0.7635782747603834,
+      "grad_norm": 3.0715521227615756,
+      "learning_rate": 9.300394538053395e-06,
+      "loss": 1.2486,
+      "step": 239
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 3.600919886468329,
+      "learning_rate": 9.29088129960862e-06,
+      "loss": 1.1424,
+      "step": 240
+    },
+    {
+      "epoch": 0.7699680511182109,
+      "grad_norm": 5.348271044142887,
+      "learning_rate": 9.281308750562426e-06,
+      "loss": 1.242,
+      "step": 241
+    },
+    {
+      "epoch": 0.7731629392971247,
+      "grad_norm": 3.265608574992943,
+      "learning_rate": 9.271677023231137e-06,
+      "loss": 1.2362,
+      "step": 242
+    },
+    {
+      "epoch": 0.7763578274760383,
+      "grad_norm": 3.439592276472071,
+      "learning_rate": 9.261986250749068e-06,
+      "loss": 1.1511,
+      "step": 243
+    },
+    {
+      "epoch": 0.7795527156549521,
+      "grad_norm": 3.6720199812303393,
+      "learning_rate": 9.252236567066686e-06,
+      "loss": 0.9605,
+      "step": 244
+    },
+    {
+      "epoch": 0.7827476038338658,
+      "grad_norm": 3.6188985353231504,
+      "learning_rate": 9.242428106948748e-06,
+      "loss": 1.1044,
+      "step": 245
+    },
+    {
+      "epoch": 0.7859424920127795,
+      "grad_norm": 3.380908179334049,
+      "learning_rate": 9.23256100597246e-06,
+      "loss": 1.0977,
+      "step": 246
+    },
+    {
+      "epoch": 0.7891373801916933,
+      "grad_norm": 3.4108419491156834,
+      "learning_rate": 9.22263540052558e-06,
+      "loss": 1.1849,
+      "step": 247
+    },
+    {
+      "epoch": 0.792332268370607,
+      "grad_norm": 3.9086904999693797,
+      "learning_rate": 9.212651427804544e-06,
+      "loss": 1.1432,
+      "step": 248
+    },
+    {
+      "epoch": 0.7955271565495208,
+      "grad_norm": 4.195148386593368,
+      "learning_rate": 9.202609225812572e-06,
+      "loss": 1.2289,
+      "step": 249
+    },
+    {
+      "epoch": 0.7987220447284346,
+      "grad_norm": 3.80521442398354,
+      "learning_rate": 9.192508933357753e-06,
+      "loss": 1.0207,
+      "step": 250
+    },
+    {
+      "epoch": 0.8019169329073482,
+      "grad_norm": 3.6843782511781,
+      "learning_rate": 9.182350690051134e-06,
+      "loss": 1.1667,
+      "step": 251
+    },
+    {
+      "epoch": 0.805111821086262,
+      "grad_norm": 4.083937273592278,
+      "learning_rate": 9.172134636304783e-06,
+      "loss": 1.086,
+      "step": 252
+    },
+    {
+      "epoch": 0.8083067092651757,
+      "grad_norm": 3.2018792661450197,
+      "learning_rate": 9.16186091332985e-06,
+      "loss": 1.0532,
+      "step": 253
+    },
+    {
+      "epoch": 0.8115015974440895,
+      "grad_norm": 4.413035483784044,
+      "learning_rate": 9.15152966313462e-06,
+      "loss": 1.2328,
+      "step": 254
+    },
+    {
+      "epoch": 0.8146964856230032,
+      "grad_norm": 3.3091915266759866,
+      "learning_rate": 9.141141028522544e-06,
+      "loss": 1.1186,
+      "step": 255
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 3.386393966028993,
+      "learning_rate": 9.130695153090272e-06,
+      "loss": 1.3441,
+      "step": 256
+    },
+    {
+      "epoch": 0.8210862619808307,
+      "grad_norm": 3.7668642876240637,
+      "learning_rate": 9.120192181225658e-06,
+      "loss": 1.1736,
+      "step": 257
+    },
+    {
+      "epoch": 0.8242811501597445,
+      "grad_norm": 3.5396964528198938,
+      "learning_rate": 9.109632258105771e-06,
+      "loss": 1.2461,
+      "step": 258
+    },
+    {
+      "epoch": 0.8274760383386581,
+      "grad_norm": 3.036695214436398,
+      "learning_rate": 9.099015529694894e-06,
+      "loss": 1.0878,
+      "step": 259
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 3.7124327249827282,
+      "learning_rate": 9.088342142742493e-06,
+      "loss": 1.25,
+      "step": 260
+    },
+    {
+      "epoch": 0.8338658146964856,
+      "grad_norm": 4.094141699956428,
+      "learning_rate": 9.077612244781196e-06,
+      "loss": 1.0391,
+      "step": 261
+    },
+    {
+      "epoch": 0.8370607028753994,
+      "grad_norm": 3.7782927364717565,
+      "learning_rate": 9.066825984124751e-06,
+      "loss": 1.1855,
+      "step": 262
+    },
+    {
+      "epoch": 0.8402555910543131,
+      "grad_norm": 3.6165644895113043,
+      "learning_rate": 9.055983509865988e-06,
+      "loss": 1.1347,
+      "step": 263
+    },
+    {
+      "epoch": 0.8434504792332268,
+      "grad_norm": 3.9591778798646273,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 1.1894,
+      "step": 264
+    },
+    {
+      "epoch": 0.8466453674121406,
+      "grad_norm": 3.0420273151626294,
+      "learning_rate": 9.034130520795774e-06,
+      "loss": 0.9918,
+      "step": 265
+    },
+    {
+      "epoch": 0.8498402555910544,
+      "grad_norm": 3.4495056799615966,
+      "learning_rate": 9.023120308046726e-06,
+      "loss": 1.1979,
+      "step": 266
+    },
+    {
+      "epoch": 0.853035143769968,
+      "grad_norm": 3.7057027439588754,
+      "learning_rate": 9.012054485815995e-06,
+      "loss": 1.0974,
+      "step": 267
+    },
+    {
+      "epoch": 0.8562300319488818,
+      "grad_norm": 4.476401090524992,
+      "learning_rate": 9.00093320706063e-06,
+      "loss": 1.2595,
+      "step": 268
+    },
+    {
+      "epoch": 0.8594249201277955,
+      "grad_norm": 3.6162896139510132,
+      "learning_rate": 8.989756625504237e-06,
+      "loss": 1.1024,
+      "step": 269
+    },
+    {
+      "epoch": 0.8626198083067093,
+      "grad_norm": 3.291618767296372,
+      "learning_rate": 8.978524895634842e-06,
+      "loss": 1.1254,
+      "step": 270
+    },
+    {
+      "epoch": 0.865814696485623,
+      "grad_norm": 3.582058254499049,
+      "learning_rate": 8.967238172702754e-06,
+      "loss": 1.3126,
+      "step": 271
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 3.364875825014213,
+      "learning_rate": 8.95589661271842e-06,
+      "loss": 1.1206,
+      "step": 272
+    },
+    {
+      "epoch": 0.8722044728434505,
+      "grad_norm": 3.482159209607752,
+      "learning_rate": 8.94450037245028e-06,
+      "loss": 1.2102,
+      "step": 273
+    },
+    {
+      "epoch": 0.8753993610223643,
+      "grad_norm": 4.565183289764591,
+      "learning_rate": 8.933049609422582e-06,
+      "loss": 1.0993,
+      "step": 274
+    },
+    {
+      "epoch": 0.8785942492012779,
+      "grad_norm": 3.2726164751259246,
+      "learning_rate": 8.921544481913218e-06,
+      "loss": 1.0898,
+      "step": 275
+    },
+    {
+      "epoch": 0.8817891373801917,
+      "grad_norm": 3.340863561728158,
+      "learning_rate": 8.909985148951528e-06,
+      "loss": 1.1027,
+      "step": 276
+    },
+    {
+      "epoch": 0.8849840255591054,
+      "grad_norm": 3.8936245184712353,
+      "learning_rate": 8.898371770316113e-06,
+      "loss": 1.3329,
+      "step": 277
+    },
+    {
+      "epoch": 0.8881789137380192,
+      "grad_norm": 3.8365956101898346,
+      "learning_rate": 8.886704506532611e-06,
+      "loss": 1.2275,
+      "step": 278
+    },
+    {
+      "epoch": 0.8913738019169329,
+      "grad_norm": 3.475763872069281,
+      "learning_rate": 8.874983518871488e-06,
+      "loss": 1.19,
+      "step": 279
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 3.7883283841356854,
+      "learning_rate": 8.86320896934581e-06,
+      "loss": 1.3191,
+      "step": 280
+    },
+    {
+      "epoch": 0.8977635782747604,
+      "grad_norm": 3.6662874802033594,
+      "learning_rate": 8.851381020709e-06,
+      "loss": 1.0672,
+      "step": 281
+    },
+    {
+      "epoch": 0.9009584664536742,
+      "grad_norm": 3.6491221926468578,
+      "learning_rate": 8.839499836452584e-06,
+      "loss": 1.0848,
+      "step": 282
+    },
+    {
+      "epoch": 0.9041533546325878,
+      "grad_norm": 4.247795106529764,
+      "learning_rate": 8.827565580803944e-06,
+      "loss": 1.0514,
+      "step": 283
+    },
+    {
+      "epoch": 0.9073482428115016,
+      "grad_norm": 3.3634695059886885,
+      "learning_rate": 8.815578418724031e-06,
+      "loss": 1.2372,
+      "step": 284
+    },
+    {
+      "epoch": 0.9105431309904153,
+      "grad_norm": 3.55364202106385,
+      "learning_rate": 8.803538515905102e-06,
+      "loss": 1.1164,
+      "step": 285
+    },
+    {
+      "epoch": 0.9137380191693291,
+      "grad_norm": 3.9706671688125414,
+      "learning_rate": 8.791446038768416e-06,
+      "loss": 1.067,
+      "step": 286
+    },
+    {
+      "epoch": 0.9169329073482428,
+      "grad_norm": 3.578148664057267,
+      "learning_rate": 8.779301154461945e-06,
+      "loss": 1.3324,
+      "step": 287
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 3.963640187538948,
+      "learning_rate": 8.76710403085805e-06,
+      "loss": 1.2401,
+      "step": 288
+    },
+    {
+      "epoch": 0.9233226837060703,
+      "grad_norm": 3.3033709430617204,
+      "learning_rate": 8.754854836551174e-06,
+      "loss": 1.0559,
+      "step": 289
+    },
+    {
+      "epoch": 0.9265175718849841,
+      "grad_norm": 3.829811290003317,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 1.1409,
+      "step": 290
+    },
+    {
+      "epoch": 0.9297124600638977,
+      "grad_norm": 3.5987381011521093,
+      "learning_rate": 8.730200913802638e-06,
+      "loss": 1.1043,
+      "step": 291
+    },
+    {
+      "epoch": 0.9329073482428115,
+      "grad_norm": 3.3931702849497523,
+      "learning_rate": 8.717796526139218e-06,
+      "loss": 1.1167,
+      "step": 292
+    },
+    {
+      "epoch": 0.9361022364217252,
+      "grad_norm": 3.403390480520602,
+      "learning_rate": 8.70534074932459e-06,
+      "loss": 1.3,
+      "step": 293
+    },
+    {
+      "epoch": 0.939297124600639,
+      "grad_norm": 3.5236398024046354,
+      "learning_rate": 8.692833755528426e-06,
+      "loss": 1.2258,
+      "step": 294
+    },
+    {
+      "epoch": 0.9424920127795527,
+      "grad_norm": 4.156604209978096,
+      "learning_rate": 8.680275717628336e-06,
+      "loss": 0.9972,
+      "step": 295
+    },
+    {
+      "epoch": 0.9456869009584664,
+      "grad_norm": 4.424919974115837,
+      "learning_rate": 8.667666809207495e-06,
+      "loss": 1.0796,
+      "step": 296
+    },
+    {
+      "epoch": 0.9488817891373802,
+      "grad_norm": 3.8179527453090905,
+      "learning_rate": 8.655007204552228e-06,
+      "loss": 1.0531,
+      "step": 297
+    },
+    {
+      "epoch": 0.952076677316294,
+      "grad_norm": 3.516161800412219,
+      "learning_rate": 8.64229707864961e-06,
+      "loss": 1.2038,
+      "step": 298
+    },
+    {
+      "epoch": 0.9552715654952076,
+      "grad_norm": 3.3495783873159257,
+      "learning_rate": 8.629536607185042e-06,
+      "loss": 1.04,
+      "step": 299
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 3.5061031729048406,
+      "learning_rate": 8.616725966539831e-06,
+      "loss": 1.153,
+      "step": 300
+    },
+    {
+      "epoch": 0.9616613418530351,
+      "grad_norm": 3.7880694893574813,
+      "learning_rate": 8.60386533378874e-06,
+      "loss": 1.2182,
+      "step": 301
+    },
+    {
+      "epoch": 0.9648562300319489,
+      "grad_norm": 3.7797890262465037,
+      "learning_rate": 8.590954886697554e-06,
+      "loss": 1.1623,
+      "step": 302
+    },
+    {
+      "epoch": 0.9680511182108626,
+      "grad_norm": 3.8141558868672485,
+      "learning_rate": 8.577994803720605e-06,
+      "loss": 1.2962,
+      "step": 303
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 3.6721022716887366,
+      "learning_rate": 8.564985263998327e-06,
+      "loss": 1.4368,
+      "step": 304
+    },
+    {
+      "epoch": 0.9744408945686901,
+      "grad_norm": 3.759951820206786,
+      "learning_rate": 8.551926447354759e-06,
+      "loss": 1.0038,
+      "step": 305
+    },
+    {
+      "epoch": 0.9776357827476039,
+      "grad_norm": 3.982402638818098,
+      "learning_rate": 8.538818534295076e-06,
+      "loss": 1.1399,
+      "step": 306
+    },
+    {
+      "epoch": 0.9808306709265175,
+      "grad_norm": 3.7009556333369513,
+      "learning_rate": 8.525661706003083e-06,
+      "loss": 1.3178,
+      "step": 307
+    },
+    {
+      "epoch": 0.9840255591054313,
+      "grad_norm": 3.1067024189717283,
+      "learning_rate": 8.512456144338717e-06,
+      "loss": 1.2427,
+      "step": 308
+    },
+    {
+      "epoch": 0.987220447284345,
+      "grad_norm": 3.624333081272346,
+      "learning_rate": 8.499202031835532e-06,
+      "loss": 1.2072,
+      "step": 309
+    },
+    {
+      "epoch": 0.9904153354632588,
+      "grad_norm": 3.894267609125984,
+      "learning_rate": 8.485899551698166e-06,
+      "loss": 1.0289,
+      "step": 310
+    },
+    {
+      "epoch": 0.9936102236421726,
+      "grad_norm": 3.3425534431748827,
+      "learning_rate": 8.472548887799833e-06,
+      "loss": 0.897,
+      "step": 311
+    },
+    {
+      "epoch": 0.9968051118210862,
+      "grad_norm": 3.9763799010157777,
+      "learning_rate": 8.45915022467975e-06,
+      "loss": 1.1306,
+      "step": 312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7070374788878517,
+      "learning_rate": 8.445703747540614e-06,
+      "loss": 1.1187,
+      "step": 313
+    },
+    {
+      "epoch": 1.0031948881789137,
+      "grad_norm": 3.3218232915565653,
+      "learning_rate": 8.43220964224602e-06,
+      "loss": 0.8005,
+      "step": 314
+    },
+    {
+      "epoch": 1.0063897763578276,
+      "grad_norm": 3.17803763928905,
+      "learning_rate": 8.418668095317912e-06,
+      "loss": 0.7299,
+      "step": 315
+    },
+    {
+      "epoch": 1.0095846645367412,
+      "grad_norm": 3.362895025540409,
+      "learning_rate": 8.405079293933986e-06,
+      "loss": 0.622,
+      "step": 316
+    },
+    {
+      "epoch": 1.012779552715655,
+      "grad_norm": 3.293744145731459,
+      "learning_rate": 8.391443425925118e-06,
+      "loss": 0.8165,
+      "step": 317
+    },
+    {
+      "epoch": 1.0159744408945688,
+      "grad_norm": 3.072987574904314,
+      "learning_rate": 8.37776067977276e-06,
+      "loss": 0.6032,
+      "step": 318
+    },
+    {
+      "epoch": 1.0191693290734825,
+      "grad_norm": 2.5714083506961107,
+      "learning_rate": 8.36403124460633e-06,
+      "loss": 0.7131,
+      "step": 319
+    },
+    {
+      "epoch": 1.0223642172523961,
+      "grad_norm": 3.463048818636885,
+      "learning_rate": 8.350255310200611e-06,
+      "loss": 0.7388,
+      "step": 320
+    },
+    {
+      "epoch": 1.0255591054313098,
+      "grad_norm": 2.9549311412515764,
+      "learning_rate": 8.336433066973122e-06,
+      "loss": 0.7718,
+      "step": 321
+    },
+    {
+      "epoch": 1.0287539936102237,
+      "grad_norm": 2.9956616912145577,
+      "learning_rate": 8.322564705981476e-06,
+      "loss": 0.6939,
+      "step": 322
+    },
+    {
+      "epoch": 1.0319488817891374,
+      "grad_norm": 3.227841274581535,
+      "learning_rate": 8.308650418920751e-06,
+      "loss": 0.7426,
+      "step": 323
+    },
+    {
+      "epoch": 1.035143769968051,
+      "grad_norm": 3.18278485706304,
+      "learning_rate": 8.294690398120843e-06,
+      "loss": 0.9084,
+      "step": 324
+    },
+    {
+      "epoch": 1.038338658146965,
+      "grad_norm": 3.4074461312131175,
+      "learning_rate": 8.280684836543794e-06,
+      "loss": 0.8289,
+      "step": 325
+    },
+    {
+      "epoch": 1.0415335463258786,
+      "grad_norm": 2.804128120870997,
+      "learning_rate": 8.266633927781135e-06,
+      "loss": 0.703,
+      "step": 326
+    },
+    {
+      "epoch": 1.0447284345047922,
+      "grad_norm": 3.5656273361765236,
+      "learning_rate": 8.25253786605121e-06,
+      "loss": 0.7009,
+      "step": 327
+    },
+    {
+      "epoch": 1.0479233226837061,
+      "grad_norm": 2.5626179881170934,
+      "learning_rate": 8.238396846196483e-06,
+      "loss": 0.5539,
+      "step": 328
+    },
+    {
+      "epoch": 1.0511182108626198,
+      "grad_norm": 3.412377464254666,
+      "learning_rate": 8.224211063680854e-06,
+      "loss": 0.9035,
+      "step": 329
+    },
+    {
+      "epoch": 1.0543130990415335,
+      "grad_norm": 2.5755297329441365,
+      "learning_rate": 8.209980714586955e-06,
+      "loss": 0.6654,
+      "step": 330
+    },
+    {
+      "epoch": 1.0575079872204474,
+      "grad_norm": 3.3218619292686657,
+      "learning_rate": 8.195705995613436e-06,
+      "loss": 0.7735,
+      "step": 331
+    },
+    {
+      "epoch": 1.060702875399361,
+      "grad_norm": 2.9112402554011734,
+      "learning_rate": 8.181387104072252e-06,
+      "loss": 0.7337,
+      "step": 332
+    },
+    {
+      "epoch": 1.0638977635782747,
+      "grad_norm": 3.41163210981966,
+      "learning_rate": 8.167024237885927e-06,
+      "loss": 0.6765,
+      "step": 333
+    },
+    {
+      "epoch": 1.0670926517571886,
+      "grad_norm": 2.9006577808675855,
+      "learning_rate": 8.152617595584827e-06,
+      "loss": 0.6437,
+      "step": 334
+    },
+    {
+      "epoch": 1.0702875399361023,
+      "grad_norm": 5.301101171489527,
+      "learning_rate": 8.138167376304411e-06,
+      "loss": 0.5427,
+      "step": 335
+    },
+    {
+      "epoch": 1.073482428115016,
+      "grad_norm": 2.6840150819679276,
+      "learning_rate": 8.123673779782481e-06,
+      "loss": 0.7049,
+      "step": 336
+    },
+    {
+      "epoch": 1.0766773162939298,
+      "grad_norm": 3.3382262094005193,
+      "learning_rate": 8.10913700635642e-06,
+      "loss": 0.6666,
+      "step": 337
+    },
+    {
+      "epoch": 1.0798722044728435,
+      "grad_norm": 3.260839917101452,
+      "learning_rate": 8.094557256960419e-06,
+      "loss": 0.6397,
+      "step": 338
+    },
+    {
+      "epoch": 1.0830670926517572,
+      "grad_norm": 3.517349038141955,
+      "learning_rate": 8.079934733122708e-06,
+      "loss": 0.5905,
+      "step": 339
+    },
+    {
+      "epoch": 1.0862619808306708,
+      "grad_norm": 3.006570919851537,
+      "learning_rate": 8.065269636962765e-06,
+      "loss": 0.6658,
+      "step": 340
+    },
+    {
+      "epoch": 1.0894568690095847,
+      "grad_norm": 3.668837476790465,
+      "learning_rate": 8.05056217118852e-06,
+      "loss": 0.6354,
+      "step": 341
+    },
+    {
+      "epoch": 1.0926517571884984,
+      "grad_norm": 3.44989817377001,
+      "learning_rate": 8.035812539093557e-06,
+      "loss": 0.8248,
+      "step": 342
+    },
+    {
+      "epoch": 1.095846645367412,
+      "grad_norm": 3.341935381892632,
+      "learning_rate": 8.021020944554305e-06,
+      "loss": 0.5529,
+      "step": 343
+    },
+    {
+      "epoch": 1.099041533546326,
+      "grad_norm": 2.9047898083709893,
+      "learning_rate": 8.006187592027215e-06,
+      "loss": 0.6516,
+      "step": 344
+    },
+    {
+      "epoch": 1.1022364217252396,
+      "grad_norm": 3.793460741106006,
+      "learning_rate": 7.991312686545939e-06,
+      "loss": 0.7692,
+      "step": 345
+    },
+    {
+      "epoch": 1.1054313099041533,
+      "grad_norm": 3.5119595706860194,
+      "learning_rate": 7.976396433718492e-06,
+      "loss": 0.7307,
+      "step": 346
+    },
+    {
+      "epoch": 1.1086261980830672,
+      "grad_norm": 4.767723072796747,
+      "learning_rate": 7.961439039724413e-06,
+      "loss": 0.6693,
+      "step": 347
+    },
+    {
+      "epoch": 1.1118210862619808,
+      "grad_norm": 3.800716857435061,
+      "learning_rate": 7.946440711311913e-06,
+      "loss": 0.5515,
+      "step": 348
+    },
+    {
+      "epoch": 1.1150159744408945,
+      "grad_norm": 2.871911978046645,
+      "learning_rate": 7.931401655795021e-06,
+      "loss": 0.6862,
+      "step": 349
+    },
+    {
+      "epoch": 1.1182108626198084,
+      "grad_norm": 3.3380609804691566,
+      "learning_rate": 7.916322081050708e-06,
+      "loss": 0.6774,
+      "step": 350
+    },
+    {
+      "epoch": 1.121405750798722,
+      "grad_norm": 3.1179289986011,
+      "learning_rate": 7.90120219551603e-06,
+      "loss": 0.8025,
+      "step": 351
+    },
+    {
+      "epoch": 1.1246006389776357,
+      "grad_norm": 2.878584569588551,
+      "learning_rate": 7.88604220818523e-06,
+      "loss": 0.6173,
+      "step": 352
+    },
+    {
+      "epoch": 1.1277955271565494,
+      "grad_norm": 3.1263915691002024,
+      "learning_rate": 7.870842328606863e-06,
+      "loss": 0.5363,
+      "step": 353
+    },
+    {
+      "epoch": 1.1309904153354633,
+      "grad_norm": 3.832199226239101,
+      "learning_rate": 7.85560276688089e-06,
+      "loss": 0.7044,
+      "step": 354
+    },
+    {
+      "epoch": 1.134185303514377,
+      "grad_norm": 3.13211968493405,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.7346,
+      "step": 355
+    },
+    {
+      "epoch": 1.1373801916932909,
+      "grad_norm": 3.7610381490543703,
+      "learning_rate": 7.825005440125595e-06,
+      "loss": 0.8411,
+      "step": 356
+    },
+    {
+      "epoch": 1.1405750798722045,
+      "grad_norm": 2.9454371449908376,
+      "learning_rate": 7.809648098027067e-06,
+      "loss": 0.5827,
+      "step": 357
+    },
+    {
+      "epoch": 1.1437699680511182,
+      "grad_norm": 3.0904272450977808,
+      "learning_rate": 7.794251919636687e-06,
+      "loss": 0.6992,
+      "step": 358
+    },
+    {
+      "epoch": 1.1469648562300319,
+      "grad_norm": 2.8056920099268416,
+      "learning_rate": 7.778817117767748e-06,
+      "loss": 0.759,
+      "step": 359
+    },
+    {
+      "epoch": 1.1501597444089458,
+      "grad_norm": 4.036274340108748,
+      "learning_rate": 7.76334390576742e-06,
+      "loss": 0.7768,
+      "step": 360
+    },
+    {
+      "epoch": 1.1533546325878594,
+      "grad_norm": 3.616745468248758,
+      "learning_rate": 7.747832497513797e-06,
+      "loss": 0.6608,
+      "step": 361
+    },
+    {
+      "epoch": 1.156549520766773,
+      "grad_norm": 2.8089850754251438,
+      "learning_rate": 7.732283107412938e-06,
+      "loss": 0.6865,
+      "step": 362
+    },
+    {
+      "epoch": 1.159744408945687,
+      "grad_norm": 2.659136093218273,
+      "learning_rate": 7.71669595039591e-06,
+      "loss": 0.5015,
+      "step": 363
+    },
+    {
+      "epoch": 1.1629392971246006,
+      "grad_norm": 3.1358419333673635,
+      "learning_rate": 7.701071241915804e-06,
+      "loss": 0.4781,
+      "step": 364
+    },
+    {
+      "epoch": 1.1661341853035143,
+      "grad_norm": 2.858994080727228,
+      "learning_rate": 7.685409197944768e-06,
+      "loss": 0.7493,
+      "step": 365
+    },
+    {
+      "epoch": 1.1693290734824282,
+      "grad_norm": 2.885070621353051,
+      "learning_rate": 7.669710034971025e-06,
+      "loss": 0.5235,
+      "step": 366
+    },
+    {
+      "epoch": 1.1725239616613419,
+      "grad_norm": 4.892494850326902,
+      "learning_rate": 7.653973969995866e-06,
+      "loss": 0.7968,
+      "step": 367
+    },
+    {
+      "epoch": 1.1757188498402555,
+      "grad_norm": 2.9392733205611523,
+      "learning_rate": 7.638201220530664e-06,
+      "loss": 0.6557,
+      "step": 368
+    },
+    {
+      "epoch": 1.1789137380191694,
+      "grad_norm": 3.8062857301355115,
+      "learning_rate": 7.622392004593862e-06,
+      "loss": 0.7582,
+      "step": 369
+    },
+    {
+      "epoch": 1.182108626198083,
+      "grad_norm": 3.5532592391497078,
+      "learning_rate": 7.60654654070796e-06,
+      "loss": 0.7568,
+      "step": 370
+    },
+    {
+      "epoch": 1.1853035143769968,
+      "grad_norm": 3.32842688263053,
+      "learning_rate": 7.59066504789649e-06,
+      "loss": 0.6446,
+      "step": 371
+    },
+    {
+      "epoch": 1.1884984025559104,
+      "grad_norm": 3.1045250925573553,
+      "learning_rate": 7.574747745681e-06,
+      "loss": 0.7385,
+      "step": 372
+    },
+    {
+      "epoch": 1.1916932907348243,
+      "grad_norm": 3.3146707460638263,
+      "learning_rate": 7.558794854078006e-06,
+      "loss": 0.59,
+      "step": 373
+    },
+    {
+      "epoch": 1.194888178913738,
+      "grad_norm": 3.283328506096695,
+      "learning_rate": 7.542806593595963e-06,
+      "loss": 0.6384,
+      "step": 374
+    },
+    {
+      "epoch": 1.1980830670926517,
+      "grad_norm": 2.603677460377522,
+      "learning_rate": 7.526783185232208e-06,
+      "loss": 0.6195,
+      "step": 375
+    },
+    {
+      "epoch": 1.2012779552715656,
+      "grad_norm": 3.209079343328633,
+      "learning_rate": 7.51072485046991e-06,
+      "loss": 0.5739,
+      "step": 376
+    },
+    {
+      "epoch": 1.2044728434504792,
+      "grad_norm": 2.8338656065965435,
+      "learning_rate": 7.494631811275008e-06,
+      "loss": 0.6693,
+      "step": 377
+    },
+    {
+      "epoch": 1.207667731629393,
+      "grad_norm": 2.8309256540806595,
+      "learning_rate": 7.478504290093138e-06,
+      "loss": 0.6675,
+      "step": 378
+    },
+    {
+      "epoch": 1.2108626198083068,
+      "grad_norm": 3.160357161294116,
+      "learning_rate": 7.462342509846571e-06,
+      "loss": 0.6601,
+      "step": 379
+    },
+    {
+      "epoch": 1.2140575079872205,
+      "grad_norm": 3.136151299398825,
+      "learning_rate": 7.446146693931111e-06,
+      "loss": 0.5592,
+      "step": 380
+    },
+    {
+      "epoch": 1.2172523961661341,
+      "grad_norm": 3.5348901357111258,
+      "learning_rate": 7.42991706621303e-06,
+      "loss": 0.7507,
+      "step": 381
+    },
+    {
+      "epoch": 1.220447284345048,
+      "grad_norm": 3.349577258862778,
+      "learning_rate": 7.413653851025959e-06,
+      "loss": 0.7736,
+      "step": 382
+    },
+    {
+      "epoch": 1.2236421725239617,
+      "grad_norm": 3.951633826678173,
+      "learning_rate": 7.397357273167789e-06,
+      "loss": 0.7067,
+      "step": 383
+    },
+    {
+      "epoch": 1.2268370607028753,
+      "grad_norm": 2.9871898551896274,
+      "learning_rate": 7.381027557897568e-06,
+      "loss": 0.677,
+      "step": 384
+    },
+    {
+      "epoch": 1.230031948881789,
+      "grad_norm": 3.012785000858288,
+      "learning_rate": 7.364664930932385e-06,
+      "loss": 0.8295,
+      "step": 385
+    },
+    {
+      "epoch": 1.233226837060703,
+      "grad_norm": 3.2999694485222926,
+      "learning_rate": 7.348269618444248e-06,
+      "loss": 0.6347,
+      "step": 386
+    },
+    {
+      "epoch": 1.2364217252396166,
+      "grad_norm": 3.07481821202855,
+      "learning_rate": 7.331841847056962e-06,
+      "loss": 0.5623,
+      "step": 387
+    },
+    {
+      "epoch": 1.2396166134185305,
+      "grad_norm": 3.4920061661357433,
+      "learning_rate": 7.315381843842995e-06,
+      "loss": 0.6764,
+      "step": 388
+    },
+    {
+      "epoch": 1.2428115015974441,
+      "grad_norm": 3.260328356860021,
+      "learning_rate": 7.298889836320334e-06,
+      "loss": 0.7761,
+      "step": 389
+    },
+    {
+      "epoch": 1.2460063897763578,
+      "grad_norm": 3.525962853972259,
+      "learning_rate": 7.282366052449351e-06,
+      "loss": 0.6741,
+      "step": 390
+    },
+    {
+      "epoch": 1.2492012779552715,
+      "grad_norm": 3.3290098348214867,
+      "learning_rate": 7.265810720629643e-06,
+      "loss": 0.6714,
+      "step": 391
+    },
+    {
+      "epoch": 1.2523961661341854,
+      "grad_norm": 3.8953250908289663,
+      "learning_rate": 7.249224069696876e-06,
+      "loss": 0.7505,
+      "step": 392
+    },
+    {
+      "epoch": 1.255591054313099,
+      "grad_norm": 3.119232906976046,
+      "learning_rate": 7.232606328919627e-06,
+      "loss": 0.6177,
+      "step": 393
+    },
+    {
+      "epoch": 1.2587859424920127,
+      "grad_norm": 3.4401928838698526,
+      "learning_rate": 7.215957727996208e-06,
+      "loss": 0.6269,
+      "step": 394
+    },
+    {
+      "epoch": 1.2619808306709266,
+      "grad_norm": 3.54623753831261,
+      "learning_rate": 7.199278497051498e-06,
+      "loss": 0.5895,
+      "step": 395
+    },
+    {
+      "epoch": 1.2651757188498403,
+      "grad_norm": 3.328882401253036,
+      "learning_rate": 7.182568866633757e-06,
+      "loss": 0.6014,
+      "step": 396
+    },
+    {
+      "epoch": 1.268370607028754,
+      "grad_norm": 3.0792693333194743,
+      "learning_rate": 7.16582906771144e-06,
+      "loss": 0.485,
+      "step": 397
+    },
+    {
+      "epoch": 1.2715654952076676,
+      "grad_norm": 3.27106754130344,
+      "learning_rate": 7.149059331670009e-06,
+      "loss": 0.8739,
+      "step": 398
+    },
+    {
+      "epoch": 1.2747603833865815,
+      "grad_norm": 3.172951547836826,
+      "learning_rate": 7.132259890308726e-06,
+      "loss": 0.6214,
+      "step": 399
+    },
+    {
+      "epoch": 1.2779552715654952,
+      "grad_norm": 2.8572122185876148,
+      "learning_rate": 7.115430975837457e-06,
+      "loss": 0.5974,
+      "step": 400
+    },
+    {
+      "epoch": 1.281150159744409,
+      "grad_norm": 2.798620093663645,
+      "learning_rate": 7.098572820873461e-06,
+      "loss": 0.631,
+      "step": 401
+    },
+    {
+      "epoch": 1.2843450479233227,
+      "grad_norm": 3.5917526129127286,
+      "learning_rate": 7.081685658438173e-06,
+      "loss": 0.8729,
+      "step": 402
+    },
+    {
+      "epoch": 1.2875399361022364,
+      "grad_norm": 2.821133789643931,
+      "learning_rate": 7.064769721953975e-06,
+      "loss": 0.6572,
+      "step": 403
+    },
+    {
+      "epoch": 1.29073482428115,
+      "grad_norm": 3.1669824440409577,
+      "learning_rate": 7.047825245240989e-06,
+      "loss": 0.8058,
+      "step": 404
+    },
+    {
+      "epoch": 1.293929712460064,
+      "grad_norm": 3.291871662213459,
+      "learning_rate": 7.030852462513827e-06,
+      "loss": 0.7929,
+      "step": 405
+    },
+    {
+      "epoch": 1.2971246006389776,
+      "grad_norm": 3.6296407682332528,
+      "learning_rate": 7.013851608378359e-06,
+      "loss": 0.6622,
+      "step": 406
+    },
+    {
+      "epoch": 1.3003194888178915,
+      "grad_norm": 4.176212377370204,
+      "learning_rate": 6.9968229178284775e-06,
+      "loss": 0.7184,
+      "step": 407
+    },
+    {
+      "epoch": 1.3035143769968052,
+      "grad_norm": 3.2838618691564223,
+      "learning_rate": 6.979766626242839e-06,
+      "loss": 0.7334,
+      "step": 408
+    },
+    {
+      "epoch": 1.3067092651757188,
+      "grad_norm": 2.948214444967057,
+      "learning_rate": 6.9626829693816135e-06,
+      "loss": 0.6316,
+      "step": 409
+    },
+    {
+      "epoch": 1.3099041533546325,
+      "grad_norm": 3.0584377089923716,
+      "learning_rate": 6.945572183383229e-06,
+      "loss": 0.6957,
+      "step": 410
+    },
+    {
+      "epoch": 1.3130990415335464,
+      "grad_norm": 3.2464259336239834,
+      "learning_rate": 6.928434504761106e-06,
+      "loss": 0.6918,
+      "step": 411
+    },
+    {
+      "epoch": 1.31629392971246,
+      "grad_norm": 3.0042894193746474,
+      "learning_rate": 6.911270170400385e-06,
+      "loss": 0.7277,
+      "step": 412
+    },
+    {
+      "epoch": 1.3194888178913737,
+      "grad_norm": 2.8761898025088812,
+      "learning_rate": 6.894079417554657e-06,
+      "loss": 0.6377,
+      "step": 413
+    },
+    {
+      "epoch": 1.3226837060702876,
+      "grad_norm": 2.9566530685808328,
+      "learning_rate": 6.8768624838426815e-06,
+      "loss": 0.5034,
+      "step": 414
+    },
+    {
+      "epoch": 1.3258785942492013,
+      "grad_norm": 3.2167209838789455,
+      "learning_rate": 6.859619607245102e-06,
+      "loss": 0.5922,
+      "step": 415
+    },
+    {
+      "epoch": 1.329073482428115,
+      "grad_norm": 3.2892505759657484,
+      "learning_rate": 6.842351026101155e-06,
+      "loss": 0.656,
+      "step": 416
+    },
+    {
+      "epoch": 1.3322683706070286,
+      "grad_norm": 3.5809700342278683,
+      "learning_rate": 6.825056979105382e-06,
+      "loss": 0.7057,
+      "step": 417
+    },
+    {
+      "epoch": 1.3354632587859425,
+      "grad_norm": 3.3755076555500247,
+      "learning_rate": 6.807737705304324e-06,
+      "loss": 0.5703,
+      "step": 418
+    },
+    {
+      "epoch": 1.3386581469648562,
+      "grad_norm": 2.7957247122701565,
+      "learning_rate": 6.790393444093214e-06,
+      "loss": 0.7173,
+      "step": 419
+    },
+    {
+      "epoch": 1.34185303514377,
+      "grad_norm": 3.6017400762275225,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.6955,
+      "step": 420
+    },
+    {
+      "epoch": 1.3450479233226837,
+      "grad_norm": 3.3729603410712974,
+      "learning_rate": 6.7556309187454185e-06,
+      "loss": 0.757,
+      "step": 421
+    },
+    {
+      "epoch": 1.3482428115015974,
+      "grad_norm": 3.150521502343688,
+      "learning_rate": 6.738213135112884e-06,
+      "loss": 0.7075,
+      "step": 422
+    },
+    {
+      "epoch": 1.351437699680511,
+      "grad_norm": 3.214955190725208,
+      "learning_rate": 6.720771325071965e-06,
+      "loss": 0.5692,
+      "step": 423
+    },
+    {
+      "epoch": 1.354632587859425,
+      "grad_norm": 3.110516238878475,
+      "learning_rate": 6.703305729711653e-06,
+      "loss": 0.6963,
+      "step": 424
+    },
+    {
+      "epoch": 1.3578274760383386,
+      "grad_norm": 3.1116100155939526,
+      "learning_rate": 6.685816590449708e-06,
+      "loss": 0.6275,
+      "step": 425
+    },
+    {
+      "epoch": 1.3610223642172525,
+      "grad_norm": 3.4423263227898184,
+      "learning_rate": 6.668304149029331e-06,
+      "loss": 0.614,
+      "step": 426
+    },
+    {
+      "epoch": 1.3642172523961662,
+      "grad_norm": 2.621180701793136,
+      "learning_rate": 6.650768647515813e-06,
+      "loss": 0.6353,
+      "step": 427
+    },
+    {
+      "epoch": 1.3674121405750799,
+      "grad_norm": 3.265293784913909,
+      "learning_rate": 6.63321032829319e-06,
+      "loss": 0.6664,
+      "step": 428
+    },
+    {
+      "epoch": 1.3706070287539935,
+      "grad_norm": 3.69932123481463,
+      "learning_rate": 6.615629434060903e-06,
+      "loss": 0.5587,
+      "step": 429
+    },
+    {
+      "epoch": 1.3738019169329074,
+      "grad_norm": 3.161206866575019,
+      "learning_rate": 6.598026207830428e-06,
+      "loss": 0.5589,
+      "step": 430
+    },
+    {
+      "epoch": 1.376996805111821,
+      "grad_norm": 3.3849192122201583,
+      "learning_rate": 6.5804008929219284e-06,
+      "loss": 0.8096,
+      "step": 431
+    },
+    {
+      "epoch": 1.3801916932907348,
+      "grad_norm": 3.4433099109238023,
+      "learning_rate": 6.562753732960887e-06,
+      "loss": 0.7941,
+      "step": 432
+    },
+    {
+      "epoch": 1.3833865814696487,
+      "grad_norm": 3.7847219844556563,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.9324,
+      "step": 433
+    },
+    {
+      "epoch": 1.3865814696485623,
+      "grad_norm": 2.8792974547251147,
+      "learning_rate": 6.527394853889499e-06,
+      "loss": 0.6378,
+      "step": 434
+    },
+    {
+      "epoch": 1.389776357827476,
+      "grad_norm": 2.820565697030664,
+      "learning_rate": 6.5096836235263904e-06,
+      "loss": 0.6333,
+      "step": 435
+    },
+    {
+      "epoch": 1.3929712460063897,
+      "grad_norm": 3.1944766223983834,
+      "learning_rate": 6.491951525598461e-06,
+      "loss": 0.5927,
+      "step": 436
+    },
+    {
+      "epoch": 1.3961661341853036,
+      "grad_norm": 3.217726988999432,
+      "learning_rate": 6.4741988052071965e-06,
+      "loss": 0.7139,
+      "step": 437
+    },
+    {
+      "epoch": 1.3993610223642172,
+      "grad_norm": 3.062225800656817,
+      "learning_rate": 6.45642570773914e-06,
+      "loss": 0.694,
+      "step": 438
+    },
+    {
+      "epoch": 1.4025559105431311,
+      "grad_norm": 3.5814717259484063,
+      "learning_rate": 6.438632478862495e-06,
+      "loss": 0.6973,
+      "step": 439
+    },
+    {
+      "epoch": 1.4057507987220448,
+      "grad_norm": 4.322441526943764,
+      "learning_rate": 6.4208193645237314e-06,
+      "loss": 0.7389,
+      "step": 440
+    },
+    {
+      "epoch": 1.4089456869009584,
+      "grad_norm": 3.469626059448132,
+      "learning_rate": 6.402986610944183e-06,
+      "loss": 0.8348,
+      "step": 441
+    },
+    {
+      "epoch": 1.4121405750798721,
+      "grad_norm": 3.1279614826433177,
+      "learning_rate": 6.385134464616649e-06,
+      "loss": 0.8252,
+      "step": 442
+    },
+    {
+      "epoch": 1.415335463258786,
+      "grad_norm": 4.087700501698058,
+      "learning_rate": 6.367263172301985e-06,
+      "loss": 0.8549,
+      "step": 443
+    },
+    {
+      "epoch": 1.4185303514376997,
+      "grad_norm": 3.241418405666963,
+      "learning_rate": 6.3493729810256895e-06,
+      "loss": 0.4999,
+      "step": 444
+    },
+    {
+      "epoch": 1.4217252396166133,
+      "grad_norm": 3.274351885074703,
+      "learning_rate": 6.331464138074493e-06,
+      "loss": 0.7477,
+      "step": 445
+    },
+    {
+      "epoch": 1.4249201277955272,
+      "grad_norm": 3.139945405727823,
+      "learning_rate": 6.313536890992935e-06,
+      "loss": 0.5821,
+      "step": 446
+    },
+    {
+      "epoch": 1.428115015974441,
+      "grad_norm": 3.337506647503942,
+      "learning_rate": 6.29559148757995e-06,
+      "loss": 0.8804,
+      "step": 447
+    },
+    {
+      "epoch": 1.4313099041533546,
+      "grad_norm": 3.240453808661668,
+      "learning_rate": 6.277628175885437e-06,
+      "loss": 0.694,
+      "step": 448
+    },
+    {
+      "epoch": 1.4345047923322682,
+      "grad_norm": 3.059498973248388,
+      "learning_rate": 6.2596472042068275e-06,
+      "loss": 0.6469,
+      "step": 449
+    },
+    {
+      "epoch": 1.4376996805111821,
+      "grad_norm": 3.300176797172696,
+      "learning_rate": 6.241648821085666e-06,
+      "loss": 0.7376,
+      "step": 450
+    },
+    {
+      "epoch": 1.4408945686900958,
+      "grad_norm": 2.7538151766754515,
+      "learning_rate": 6.223633275304157e-06,
+      "loss": 0.6763,
+      "step": 451
+    },
+    {
+      "epoch": 1.4440894568690097,
+      "grad_norm": 3.0614989291769783,
+      "learning_rate": 6.205600815881741e-06,
+      "loss": 0.7564,
+      "step": 452
+    },
+    {
+      "epoch": 1.4472843450479234,
+      "grad_norm": 3.674154196780988,
+      "learning_rate": 6.187551692071648e-06,
+      "loss": 0.743,
+      "step": 453
+    },
+    {
+      "epoch": 1.450479233226837,
+      "grad_norm": 3.601624639870162,
+      "learning_rate": 6.1694861533574445e-06,
+      "loss": 0.5951,
+      "step": 454
+    },
+    {
+      "epoch": 1.4536741214057507,
+      "grad_norm": 3.222869188636421,
+      "learning_rate": 6.1514044494496e-06,
+      "loss": 0.709,
+      "step": 455
+    },
+    {
+      "epoch": 1.4568690095846646,
+      "grad_norm": 3.836073976010034,
+      "learning_rate": 6.133306830282021e-06,
+      "loss": 0.892,
+      "step": 456
+    },
+    {
+      "epoch": 1.4600638977635783,
+      "grad_norm": 3.1116863144500377,
+      "learning_rate": 6.115193546008602e-06,
+      "loss": 0.7786,
+      "step": 457
+    },
+    {
+      "epoch": 1.4632587859424921,
+      "grad_norm": 2.9960474146151017,
+      "learning_rate": 6.097064846999774e-06,
+      "loss": 0.7682,
+      "step": 458
+    },
+    {
+      "epoch": 1.4664536741214058,
+      "grad_norm": 3.55443055142297,
+      "learning_rate": 6.078920983839032e-06,
+      "loss": 0.7227,
+      "step": 459
+    },
+    {
+      "epoch": 1.4696485623003195,
+      "grad_norm": 3.80776666518291,
+      "learning_rate": 6.060762207319479e-06,
+      "loss": 0.7646,
+      "step": 460
+    },
+    {
+      "epoch": 1.4728434504792332,
+      "grad_norm": 2.9668398977788457,
+      "learning_rate": 6.042588768440358e-06,
+      "loss": 0.7479,
+      "step": 461
+    },
+    {
+      "epoch": 1.476038338658147,
+      "grad_norm": 3.751918555792093,
+      "learning_rate": 6.024400918403581e-06,
+      "loss": 0.8015,
+      "step": 462
+    },
+    {
+      "epoch": 1.4792332268370607,
+      "grad_norm": 2.9766998953540047,
+      "learning_rate": 6.006198908610261e-06,
+      "loss": 0.6295,
+      "step": 463
+    },
+    {
+      "epoch": 1.4824281150159744,
+      "grad_norm": 4.106805870394821,
+      "learning_rate": 5.987982990657229e-06,
+      "loss": 0.6329,
+      "step": 464
+    },
+    {
+      "epoch": 1.4856230031948883,
+      "grad_norm": 3.4779890506613365,
+      "learning_rate": 5.9697534163335645e-06,
+      "loss": 0.8793,
+      "step": 465
+    },
+    {
+      "epoch": 1.488817891373802,
+      "grad_norm": 3.4529058427373203,
+      "learning_rate": 5.95151043761711e-06,
+      "loss": 0.6773,
+      "step": 466
+    },
+    {
+      "epoch": 1.4920127795527156,
+      "grad_norm": 2.9478809202331355,
+      "learning_rate": 5.933254306670995e-06,
+      "loss": 0.5695,
+      "step": 467
+    },
+    {
+      "epoch": 1.4952076677316293,
+      "grad_norm": 3.216077759029646,
+      "learning_rate": 5.914985275840135e-06,
+      "loss": 0.7331,
+      "step": 468
+    },
+    {
+      "epoch": 1.4984025559105432,
+      "grad_norm": 2.7383943066727108,
+      "learning_rate": 5.896703597647765e-06,
+      "loss": 0.5549,
+      "step": 469
+    },
+    {
+      "epoch": 1.5015974440894568,
+      "grad_norm": 3.3584760866398726,
+      "learning_rate": 5.878409524791931e-06,
+      "loss": 0.7706,
+      "step": 470
+    },
+    {
+      "epoch": 1.5047923322683707,
+      "grad_norm": 3.1951595906138492,
+      "learning_rate": 5.8601033101420055e-06,
+      "loss": 0.7557,
+      "step": 471
+    },
+    {
+      "epoch": 1.5079872204472844,
+      "grad_norm": 3.3245430960441262,
+      "learning_rate": 5.841785206735192e-06,
+      "loss": 0.7702,
+      "step": 472
+    },
+    {
+      "epoch": 1.511182108626198,
+      "grad_norm": 2.7680348059241795,
+      "learning_rate": 5.823455467773027e-06,
+      "loss": 0.7489,
+      "step": 473
+    },
+    {
+      "epoch": 1.5143769968051117,
+      "grad_norm": 3.1622873952189208,
+      "learning_rate": 5.805114346617874e-06,
+      "loss": 0.8298,
+      "step": 474
+    },
+    {
+      "epoch": 1.5175718849840254,
+      "grad_norm": 3.280875590039776,
+      "learning_rate": 5.786762096789431e-06,
+      "loss": 0.5995,
+      "step": 475
+    },
+    {
+      "epoch": 1.5207667731629393,
+      "grad_norm": 3.1578595322517136,
+      "learning_rate": 5.768398971961221e-06,
+      "loss": 0.6651,
+      "step": 476
+    },
+    {
+      "epoch": 1.5239616613418532,
+      "grad_norm": 3.279478924470811,
+      "learning_rate": 5.750025225957086e-06,
+      "loss": 0.756,
+      "step": 477
+    },
+    {
+      "epoch": 1.5271565495207668,
+      "grad_norm": 3.4655787627014116,
+      "learning_rate": 5.731641112747679e-06,
+      "loss": 0.4961,
+      "step": 478
+    },
+    {
+      "epoch": 1.5303514376996805,
+      "grad_norm": 3.19538645637906,
+      "learning_rate": 5.713246886446954e-06,
+      "loss": 0.6418,
+      "step": 479
+    },
+    {
+      "epoch": 1.5335463258785942,
+      "grad_norm": 3.118191510242903,
+      "learning_rate": 5.694842801308651e-06,
+      "loss": 0.7207,
+      "step": 480
+    },
+    {
+      "epoch": 1.5367412140575079,
+      "grad_norm": 3.1367578960090925,
+      "learning_rate": 5.676429111722786e-06,
+      "loss": 0.6739,
+      "step": 481
+    },
+    {
+      "epoch": 1.5399361022364217,
+      "grad_norm": 3.1900161302085537,
+      "learning_rate": 5.6580060722121325e-06,
+      "loss": 0.6686,
+      "step": 482
+    },
+    {
+      "epoch": 1.5431309904153354,
+      "grad_norm": 3.3079346250335346,
+      "learning_rate": 5.639573937428699e-06,
+      "loss": 0.5788,
+      "step": 483
+    },
+    {
+      "epoch": 1.5463258785942493,
+      "grad_norm": 3.3321492619064594,
+      "learning_rate": 5.621132962150216e-06,
+      "loss": 0.6126,
+      "step": 484
+    },
+    {
+      "epoch": 1.549520766773163,
+      "grad_norm": 3.685843053367816,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.8318,
+      "step": 485
+    },
+    {
+      "epoch": 1.5527156549520766,
+      "grad_norm": 3.3695267494165293,
+      "learning_rate": 5.584225509826497e-06,
+      "loss": 0.5958,
+      "step": 486
+    },
+    {
+      "epoch": 1.5559105431309903,
+      "grad_norm": 3.1293118828855193,
+      "learning_rate": 5.565759542933612e-06,
+      "loss": 0.6171,
+      "step": 487
+    },
+    {
+      "epoch": 1.5591054313099042,
+      "grad_norm": 4.0342603862316215,
+      "learning_rate": 5.547285755843334e-06,
+      "loss": 0.6896,
+      "step": 488
+    },
+    {
+      "epoch": 1.5623003194888179,
+      "grad_norm": 3.1031607445308294,
+      "learning_rate": 5.5288044039091335e-06,
+      "loss": 0.7725,
+      "step": 489
+    },
+    {
+      "epoch": 1.5654952076677318,
+      "grad_norm": 2.8870551106465268,
+      "learning_rate": 5.510315742589042e-06,
+      "loss": 0.7087,
+      "step": 490
+    },
+    {
+      "epoch": 1.5686900958466454,
+      "grad_norm": 3.223157564919875,
+      "learning_rate": 5.491820027442126e-06,
+      "loss": 0.7724,
+      "step": 491
+    },
+    {
+      "epoch": 1.571884984025559,
+      "grad_norm": 4.374819989730027,
+      "learning_rate": 5.473317514124958e-06,
+      "loss": 0.9128,
+      "step": 492
+    },
+    {
+      "epoch": 1.5750798722044728,
+      "grad_norm": 3.4857978259475977,
+      "learning_rate": 5.454808458388069e-06,
+      "loss": 0.6537,
+      "step": 493
+    },
+    {
+      "epoch": 1.5782747603833864,
+      "grad_norm": 3.2306356545828816,
+      "learning_rate": 5.436293116072431e-06,
+      "loss": 0.7887,
+      "step": 494
+    },
+    {
+      "epoch": 1.5814696485623003,
+      "grad_norm": 3.1584271025921637,
+      "learning_rate": 5.417771743105908e-06,
+      "loss": 0.6372,
+      "step": 495
+    },
+    {
+      "epoch": 1.5846645367412142,
+      "grad_norm": 2.947143879561285,
+      "learning_rate": 5.399244595499721e-06,
+      "loss": 0.475,
+      "step": 496
+    },
+    {
+      "epoch": 1.5878594249201279,
+      "grad_norm": 3.0861995387113907,
+      "learning_rate": 5.380711929344915e-06,
+      "loss": 0.6427,
+      "step": 497
+    },
+    {
+      "epoch": 1.5910543130990416,
+      "grad_norm": 3.075484857090756,
+      "learning_rate": 5.362174000808813e-06,
+      "loss": 0.7657,
+      "step": 498
+    },
+    {
+      "epoch": 1.5942492012779552,
+      "grad_norm": 3.0280477925820217,
+      "learning_rate": 5.343631066131476e-06,
+      "loss": 0.507,
+      "step": 499
+    },
+    {
+      "epoch": 1.5974440894568689,
+      "grad_norm": 3.900387190666821,
+      "learning_rate": 5.325083381622165e-06,
+      "loss": 0.7399,
+      "step": 500
+    },
+    {
+      "epoch": 1.6006389776357828,
+      "grad_norm": 3.0285358655321333,
+      "learning_rate": 5.30653120365579e-06,
+      "loss": 0.7796,
+      "step": 501
+    },
+    {
+      "epoch": 1.6038338658146964,
+      "grad_norm": 3.1107733171323093,
+      "learning_rate": 5.28797478866938e-06,
+      "loss": 0.6737,
+      "step": 502
+    },
+    {
+      "epoch": 1.6070287539936103,
+      "grad_norm": 3.773613303681798,
+      "learning_rate": 5.269414393158523e-06,
+      "loss": 0.6916,
+      "step": 503
+    },
+    {
+      "epoch": 1.610223642172524,
+      "grad_norm": 3.3401282021294993,
+      "learning_rate": 5.250850273673831e-06,
+      "loss": 0.6725,
+      "step": 504
+    },
+    {
+      "epoch": 1.6134185303514377,
+      "grad_norm": 2.949197013433841,
+      "learning_rate": 5.232282686817392e-06,
+      "loss": 0.6294,
+      "step": 505
+    },
+    {
+      "epoch": 1.6166134185303513,
+      "grad_norm": 3.237588483170284,
+      "learning_rate": 5.213711889239214e-06,
+      "loss": 0.7455,
+      "step": 506
+    },
+    {
+      "epoch": 1.619808306709265,
+      "grad_norm": 3.1132611565211867,
+      "learning_rate": 5.195138137633695e-06,
+      "loss": 0.5668,
+      "step": 507
+    },
+    {
+      "epoch": 1.623003194888179,
+      "grad_norm": 3.547445257711987,
+      "learning_rate": 5.17656168873606e-06,
+      "loss": 0.6301,
+      "step": 508
+    },
+    {
+      "epoch": 1.6261980830670928,
+      "grad_norm": 3.3919535025909693,
+      "learning_rate": 5.157982799318817e-06,
+      "loss": 0.8712,
+      "step": 509
+    },
+    {
+      "epoch": 1.6293929712460065,
+      "grad_norm": 3.145942215941063,
+      "learning_rate": 5.139401726188208e-06,
+      "loss": 0.6416,
+      "step": 510
+    },
+    {
+      "epoch": 1.6325878594249201,
+      "grad_norm": 3.234750707094919,
+      "learning_rate": 5.120818726180662e-06,
+      "loss": 0.8133,
+      "step": 511
+    },
+    {
+      "epoch": 1.6357827476038338,
+      "grad_norm": 4.150262641585497,
+      "learning_rate": 5.1022340561592396e-06,
+      "loss": 0.7914,
+      "step": 512
+    },
+    {
+      "epoch": 1.6389776357827475,
+      "grad_norm": 3.131123530132958,
+      "learning_rate": 5.083647973010085e-06,
+      "loss": 0.6892,
+      "step": 513
+    },
+    {
+      "epoch": 1.6421725239616614,
+      "grad_norm": 2.994597770740181,
+      "learning_rate": 5.065060733638878e-06,
+      "loss": 0.8447,
+      "step": 514
+    },
+    {
+      "epoch": 1.645367412140575,
+      "grad_norm": 3.2479942609957453,
+      "learning_rate": 5.046472594967279e-06,
+      "loss": 0.5937,
+      "step": 515
+    },
+    {
+      "epoch": 1.648562300319489,
+      "grad_norm": 3.7109334936610354,
+      "learning_rate": 5.027883813929374e-06,
+      "loss": 0.909,
+      "step": 516
+    },
+    {
+      "epoch": 1.6517571884984026,
+      "grad_norm": 2.6524679804824944,
+      "learning_rate": 5.009294647468137e-06,
+      "loss": 0.7518,
+      "step": 517
+    },
+    {
+      "epoch": 1.6549520766773163,
+      "grad_norm": 2.838493088028963,
+      "learning_rate": 4.990705352531864e-06,
+      "loss": 0.7304,
+      "step": 518
+    },
+    {
+      "epoch": 1.65814696485623,
+      "grad_norm": 3.1776338235528114,
+      "learning_rate": 4.972116186070626e-06,
+      "loss": 0.7314,
+      "step": 519
+    },
+    {
+      "epoch": 1.6613418530351438,
+      "grad_norm": 3.426789598514184,
+      "learning_rate": 4.953527405032723e-06,
+      "loss": 0.7108,
+      "step": 520
+    },
+    {
+      "epoch": 1.6645367412140575,
+      "grad_norm": 4.858962396820881,
+      "learning_rate": 4.934939266361123e-06,
+      "loss": 0.6469,
+      "step": 521
+    },
+    {
+      "epoch": 1.6677316293929714,
+      "grad_norm": 3.5004561539172,
+      "learning_rate": 4.916352026989914e-06,
+      "loss": 0.7428,
+      "step": 522
+    },
+    {
+      "epoch": 1.670926517571885,
+      "grad_norm": 3.3312627947706175,
+      "learning_rate": 4.897765943840761e-06,
+      "loss": 0.668,
+      "step": 523
+    },
+    {
+      "epoch": 1.6741214057507987,
+      "grad_norm": 2.9986061479971173,
+      "learning_rate": 4.87918127381934e-06,
+      "loss": 0.6034,
+      "step": 524
+    },
+    {
+      "epoch": 1.6773162939297124,
+      "grad_norm": 3.3503743425109604,
+      "learning_rate": 4.860598273811793e-06,
+      "loss": 0.6615,
+      "step": 525
+    },
+    {
+      "epoch": 1.680511182108626,
+      "grad_norm": 3.3329168106048432,
+      "learning_rate": 4.842017200681185e-06,
+      "loss": 0.7828,
+      "step": 526
+    },
+    {
+      "epoch": 1.68370607028754,
+      "grad_norm": 3.689137249220177,
+      "learning_rate": 4.823438311263943e-06,
+      "loss": 0.7517,
+      "step": 527
+    },
+    {
+      "epoch": 1.6869009584664538,
+      "grad_norm": 3.9483469707540686,
+      "learning_rate": 4.804861862366306e-06,
+      "loss": 0.8853,
+      "step": 528
+    },
+    {
+      "epoch": 1.6900958466453675,
+      "grad_norm": 3.2086456717020084,
+      "learning_rate": 4.786288110760787e-06,
+      "loss": 0.6986,
+      "step": 529
+    },
+    {
+      "epoch": 1.6932907348242812,
+      "grad_norm": 3.8086620367137938,
+      "learning_rate": 4.767717313182611e-06,
+      "loss": 0.7893,
+      "step": 530
+    },
+    {
+      "epoch": 1.6964856230031948,
+      "grad_norm": 3.513673584809838,
+      "learning_rate": 4.74914972632617e-06,
+      "loss": 0.6758,
+      "step": 531
+    },
+    {
+      "epoch": 1.6996805111821085,
+      "grad_norm": 2.9173324944934733,
+      "learning_rate": 4.730585606841479e-06,
+      "loss": 0.6683,
+      "step": 532
+    },
+    {
+      "epoch": 1.7028753993610224,
+      "grad_norm": 3.6380954918455988,
+      "learning_rate": 4.7120252113306216e-06,
+      "loss": 0.8242,
+      "step": 533
+    },
+    {
+      "epoch": 1.706070287539936,
+      "grad_norm": 3.7350889701735186,
+      "learning_rate": 4.693468796344211e-06,
+      "loss": 0.7802,
+      "step": 534
+    },
+    {
+      "epoch": 1.70926517571885,
+      "grad_norm": 3.4854571190867842,
+      "learning_rate": 4.6749166183778375e-06,
+      "loss": 0.6251,
+      "step": 535
+    },
+    {
+      "epoch": 1.7124600638977636,
+      "grad_norm": 2.809387896584307,
+      "learning_rate": 4.656368933868525e-06,
+      "loss": 0.6736,
+      "step": 536
+    },
+    {
+      "epoch": 1.7156549520766773,
+      "grad_norm": 3.2807881673977906,
+      "learning_rate": 4.637825999191189e-06,
+      "loss": 0.6122,
+      "step": 537
+    },
+    {
+      "epoch": 1.718849840255591,
+      "grad_norm": 4.34914543869363,
+      "learning_rate": 4.619288070655086e-06,
+      "loss": 1.0081,
+      "step": 538
+    },
+    {
+      "epoch": 1.7220447284345048,
+      "grad_norm": 3.2664054679831547,
+      "learning_rate": 4.600755404500281e-06,
+      "loss": 0.8357,
+      "step": 539
+    },
+    {
+      "epoch": 1.7252396166134185,
+      "grad_norm": 3.0971094315408383,
+      "learning_rate": 4.582228256894093e-06,
+      "loss": 0.7042,
+      "step": 540
+    },
+    {
+      "epoch": 1.7284345047923324,
+      "grad_norm": 3.4780964078654897,
+      "learning_rate": 4.56370688392757e-06,
+      "loss": 0.7841,
+      "step": 541
+    },
+    {
+      "epoch": 1.731629392971246,
+      "grad_norm": 3.387358223123862,
+      "learning_rate": 4.545191541611933e-06,
+      "loss": 0.7654,
+      "step": 542
+    },
+    {
+      "epoch": 1.7348242811501597,
+      "grad_norm": 2.938067759114242,
+      "learning_rate": 4.526682485875044e-06,
+      "loss": 0.7947,
+      "step": 543
+    },
+    {
+      "epoch": 1.7380191693290734,
+      "grad_norm": 2.7043603615698824,
+      "learning_rate": 4.508179972557875e-06,
+      "loss": 0.7439,
+      "step": 544
+    },
+    {
+      "epoch": 1.741214057507987,
+      "grad_norm": 3.254085901019241,
+      "learning_rate": 4.489684257410959e-06,
+      "loss": 0.687,
+      "step": 545
+    },
+    {
+      "epoch": 1.744408945686901,
+      "grad_norm": 3.5128781763466828,
+      "learning_rate": 4.471195596090867e-06,
+      "loss": 0.6303,
+      "step": 546
+    },
+    {
+      "epoch": 1.7476038338658149,
+      "grad_norm": 3.320148960757458,
+      "learning_rate": 4.452714244156667e-06,
+      "loss": 0.5445,
+      "step": 547
+    },
+    {
+      "epoch": 1.7507987220447285,
+      "grad_norm": 3.509779618698661,
+      "learning_rate": 4.434240457066388e-06,
+      "loss": 0.743,
+      "step": 548
+    },
+    {
+      "epoch": 1.7539936102236422,
+      "grad_norm": 2.8821825101484593,
+      "learning_rate": 4.415774490173504e-06,
+      "loss": 0.8359,
+      "step": 549
+    },
+    {
+      "epoch": 1.7571884984025559,
+      "grad_norm": 3.615093620336388,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.8082,
+      "step": 550
+    },
+    {
+      "epoch": 1.7603833865814695,
+      "grad_norm": 3.7926346941563445,
+      "learning_rate": 4.3788670378497836e-06,
+      "loss": 0.6949,
+      "step": 551
+    },
+    {
+      "epoch": 1.7635782747603834,
+      "grad_norm": 3.340725204233956,
+      "learning_rate": 4.360426062571303e-06,
+      "loss": 0.472,
+      "step": 552
+    },
+    {
+      "epoch": 1.766773162939297,
+      "grad_norm": 3.574725506884094,
+      "learning_rate": 4.341993927787871e-06,
+      "loss": 0.6599,
+      "step": 553
+    },
+    {
+      "epoch": 1.769968051118211,
+      "grad_norm": 3.112184327281778,
+      "learning_rate": 4.323570888277215e-06,
+      "loss": 0.6121,
+      "step": 554
+    },
+    {
+      "epoch": 1.7731629392971247,
+      "grad_norm": 3.652894953819498,
+      "learning_rate": 4.305157198691351e-06,
+      "loss": 0.6701,
+      "step": 555
+    },
+    {
+      "epoch": 1.7763578274760383,
+      "grad_norm": 3.226329998839247,
+      "learning_rate": 4.286753113553049e-06,
+      "loss": 0.6362,
+      "step": 556
+    },
+    {
+      "epoch": 1.779552715654952,
+      "grad_norm": 3.4711880355889995,
+      "learning_rate": 4.268358887252322e-06,
+      "loss": 0.7042,
+      "step": 557
+    },
+    {
+      "epoch": 1.7827476038338657,
+      "grad_norm": 3.2435653923901326,
+      "learning_rate": 4.249974774042915e-06,
+      "loss": 0.6979,
+      "step": 558
+    },
+    {
+      "epoch": 1.7859424920127795,
+      "grad_norm": 2.795354227800207,
+      "learning_rate": 4.231601028038781e-06,
+      "loss": 0.7669,
+      "step": 559
+    },
+    {
+      "epoch": 1.7891373801916934,
+      "grad_norm": 3.9900946233369807,
+      "learning_rate": 4.2132379032105695e-06,
+      "loss": 0.7462,
+      "step": 560
+    },
+    {
+      "epoch": 1.792332268370607,
+      "grad_norm": 3.3921685720797874,
+      "learning_rate": 4.194885653382128e-06,
+      "loss": 0.5706,
+      "step": 561
+    },
+    {
+      "epoch": 1.7955271565495208,
+      "grad_norm": 2.9067872235604653,
+      "learning_rate": 4.176544532226974e-06,
+      "loss": 0.4336,
+      "step": 562
+    },
+    {
+      "epoch": 1.7987220447284344,
+      "grad_norm": 3.3106094389455563,
+      "learning_rate": 4.158214793264808e-06,
+      "loss": 0.5993,
+      "step": 563
+    },
+    {
+      "epoch": 1.8019169329073481,
+      "grad_norm": 3.4439326566007775,
+      "learning_rate": 4.139896689857995e-06,
+      "loss": 0.7628,
+      "step": 564
+    },
+    {
+      "epoch": 1.805111821086262,
+      "grad_norm": 3.101900195962954,
+      "learning_rate": 4.121590475208071e-06,
+      "loss": 0.7233,
+      "step": 565
+    },
+    {
+      "epoch": 1.8083067092651757,
+      "grad_norm": 3.320947068170224,
+      "learning_rate": 4.1032964023522366e-06,
+      "loss": 0.6094,
+      "step": 566
+    },
+    {
+      "epoch": 1.8115015974440896,
+      "grad_norm": 2.8470212218375943,
+      "learning_rate": 4.085014724159866e-06,
+      "loss": 0.583,
+      "step": 567
+    },
+    {
+      "epoch": 1.8146964856230032,
+      "grad_norm": 3.334412016602001,
+      "learning_rate": 4.066745693329008e-06,
+      "loss": 0.661,
+      "step": 568
+    },
+    {
+      "epoch": 1.817891373801917,
+      "grad_norm": 2.9561564236547286,
+      "learning_rate": 4.0484895623828906e-06,
+      "loss": 0.5916,
+      "step": 569
+    },
+    {
+      "epoch": 1.8210862619808306,
+      "grad_norm": 3.222543757516836,
+      "learning_rate": 4.030246583666437e-06,
+      "loss": 0.8526,
+      "step": 570
+    },
+    {
+      "epoch": 1.8242811501597445,
+      "grad_norm": 2.9069484871983584,
+      "learning_rate": 4.012017009342773e-06,
+      "loss": 0.6607,
+      "step": 571
+    },
+    {
+      "epoch": 1.8274760383386581,
+      "grad_norm": 3.371649685649377,
+      "learning_rate": 3.99380109138974e-06,
+      "loss": 0.7,
+      "step": 572
+    },
+    {
+      "epoch": 1.830670926517572,
+      "grad_norm": 3.0490496787344386,
+      "learning_rate": 3.97559908159642e-06,
+      "loss": 0.5186,
+      "step": 573
+    },
+    {
+      "epoch": 1.8338658146964857,
+      "grad_norm": 2.4634468653823887,
+      "learning_rate": 3.9574112315596425e-06,
+      "loss": 0.5184,
+      "step": 574
+    },
+    {
+      "epoch": 1.8370607028753994,
+      "grad_norm": 2.9813271614172074,
+      "learning_rate": 3.9392377926805226e-06,
+      "loss": 0.7424,
+      "step": 575
+    },
+    {
+      "epoch": 1.840255591054313,
+      "grad_norm": 3.545723916042709,
+      "learning_rate": 3.92107901616097e-06,
+      "loss": 0.6711,
+      "step": 576
+    },
+    {
+      "epoch": 1.8434504792332267,
+      "grad_norm": 3.642332037679665,
+      "learning_rate": 3.9029351530002264e-06,
+      "loss": 0.6889,
+      "step": 577
+    },
+    {
+      "epoch": 1.8466453674121406,
+      "grad_norm": 2.9694687429994695,
+      "learning_rate": 3.884806453991399e-06,
+      "loss": 0.5221,
+      "step": 578
+    },
+    {
+      "epoch": 1.8498402555910545,
+      "grad_norm": 3.771938049168385,
+      "learning_rate": 3.866693169717982e-06,
+      "loss": 0.7594,
+      "step": 579
+    },
+    {
+      "epoch": 1.8530351437699681,
+      "grad_norm": 4.044966561123037,
+      "learning_rate": 3.848595550550401e-06,
+      "loss": 0.7512,
+      "step": 580
+    },
+    {
+      "epoch": 1.8562300319488818,
+      "grad_norm": 3.382330795701972,
+      "learning_rate": 3.830513846642556e-06,
+      "loss": 0.7434,
+      "step": 581
+    },
+    {
+      "epoch": 1.8594249201277955,
+      "grad_norm": 3.18626360833024,
+      "learning_rate": 3.8124483079283546e-06,
+      "loss": 0.6854,
+      "step": 582
+    },
+    {
+      "epoch": 1.8626198083067091,
+      "grad_norm": 3.75731580169973,
+      "learning_rate": 3.7943991841182586e-06,
+      "loss": 0.7578,
+      "step": 583
+    },
+    {
+      "epoch": 1.865814696485623,
+      "grad_norm": 3.2642841240715104,
+      "learning_rate": 3.7763667246958447e-06,
+      "loss": 0.8174,
+      "step": 584
+    },
+    {
+      "epoch": 1.8690095846645367,
+      "grad_norm": 4.0936621279127205,
+      "learning_rate": 3.758351178914336e-06,
+      "loss": 0.7589,
+      "step": 585
+    },
+    {
+      "epoch": 1.8722044728434506,
+      "grad_norm": 3.733067656417201,
+      "learning_rate": 3.7403527957931716e-06,
+      "loss": 0.6164,
+      "step": 586
+    },
+    {
+      "epoch": 1.8753993610223643,
+      "grad_norm": 3.1679037152983778,
+      "learning_rate": 3.7223718241145646e-06,
+      "loss": 0.753,
+      "step": 587
+    },
+    {
+      "epoch": 1.878594249201278,
+      "grad_norm": 3.186379548695223,
+      "learning_rate": 3.7044085124200517e-06,
+      "loss": 0.5979,
+      "step": 588
+    },
+    {
+      "epoch": 1.8817891373801916,
+      "grad_norm": 3.4320599617810537,
+      "learning_rate": 3.6864631090070656e-06,
+      "loss": 0.6803,
+      "step": 589
+    },
+    {
+      "epoch": 1.8849840255591053,
+      "grad_norm": 2.842925328333586,
+      "learning_rate": 3.668535861925509e-06,
+      "loss": 0.5987,
+      "step": 590
+    },
+    {
+      "epoch": 1.8881789137380192,
+      "grad_norm": 5.945517377925381,
+      "learning_rate": 3.650627018974312e-06,
+      "loss": 0.7192,
+      "step": 591
+    },
+    {
+      "epoch": 1.891373801916933,
+      "grad_norm": 3.5110013263580933,
+      "learning_rate": 3.632736827698015e-06,
+      "loss": 0.7895,
+      "step": 592
+    },
+    {
+      "epoch": 1.8945686900958467,
+      "grad_norm": 3.417943110299541,
+      "learning_rate": 3.6148655353833518e-06,
+      "loss": 0.6726,
+      "step": 593
+    },
+    {
+      "epoch": 1.8977635782747604,
+      "grad_norm": 2.9333343934591714,
+      "learning_rate": 3.5970133890558184e-06,
+      "loss": 0.7217,
+      "step": 594
+    },
+    {
+      "epoch": 1.900958466453674,
+      "grad_norm": 4.0141130310790505,
+      "learning_rate": 3.5791806354762702e-06,
+      "loss": 0.6631,
+      "step": 595
+    },
+    {
+      "epoch": 1.9041533546325877,
+      "grad_norm": 3.2672598335362224,
+      "learning_rate": 3.5613675211375066e-06,
+      "loss": 0.6694,
+      "step": 596
+    },
+    {
+      "epoch": 1.9073482428115016,
+      "grad_norm": 2.887149036765588,
+      "learning_rate": 3.5435742922608618e-06,
+      "loss": 0.611,
+      "step": 597
+    },
+    {
+      "epoch": 1.9105431309904153,
+      "grad_norm": 3.5964941537613893,
+      "learning_rate": 3.525801194792805e-06,
+      "loss": 0.7054,
+      "step": 598
+    },
+    {
+      "epoch": 1.9137380191693292,
+      "grad_norm": 3.396197000077211,
+      "learning_rate": 3.508048474401541e-06,
+      "loss": 0.6918,
+      "step": 599
+    },
+    {
+      "epoch": 1.9169329073482428,
+      "grad_norm": 3.323762665131941,
+      "learning_rate": 3.4903163764736104e-06,
+      "loss": 0.7055,
+      "step": 600
+    },
+    {
+      "epoch": 1.9201277955271565,
+      "grad_norm": 3.3226138548254442,
+      "learning_rate": 3.4726051461105016e-06,
+      "loss": 0.7109,
+      "step": 601
+    },
+    {
+      "epoch": 1.9233226837060702,
+      "grad_norm": 3.1266458943976345,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.546,
+      "step": 602
+    },
+    {
+      "epoch": 1.926517571884984,
+      "grad_norm": 3.0946223283386867,
+      "learning_rate": 3.437246267039115e-06,
+      "loss": 0.7747,
+      "step": 603
+    },
+    {
+      "epoch": 1.9297124600638977,
+      "grad_norm": 2.750088967836726,
+      "learning_rate": 3.419599107078073e-06,
+      "loss": 0.5863,
+      "step": 604
+    },
+    {
+      "epoch": 1.9329073482428116,
+      "grad_norm": 2.514163428602213,
+      "learning_rate": 3.401973792169574e-06,
+      "loss": 0.6824,
+      "step": 605
+    },
+    {
+      "epoch": 1.9361022364217253,
+      "grad_norm": 2.96990880222507,
+      "learning_rate": 3.384370565939098e-06,
+      "loss": 0.4532,
+      "step": 606
+    },
+    {
+      "epoch": 1.939297124600639,
+      "grad_norm": 3.4153565739906724,
+      "learning_rate": 3.3667896717068105e-06,
+      "loss": 0.7031,
+      "step": 607
+    },
+    {
+      "epoch": 1.9424920127795526,
+      "grad_norm": 3.5800017509722917,
+      "learning_rate": 3.34923135248419e-06,
+      "loss": 0.5049,
+      "step": 608
+    },
+    {
+      "epoch": 1.9456869009584663,
+      "grad_norm": 4.046001931299702,
+      "learning_rate": 3.33169585097067e-06,
+      "loss": 0.5494,
+      "step": 609
+    },
+    {
+      "epoch": 1.9488817891373802,
+      "grad_norm": 2.8841226611862743,
+      "learning_rate": 3.314183409550293e-06,
+      "loss": 0.5554,
+      "step": 610
+    },
+    {
+      "epoch": 1.952076677316294,
+      "grad_norm": 3.588955989839378,
+      "learning_rate": 3.2966942702883494e-06,
+      "loss": 0.7216,
+      "step": 611
+    },
+    {
+      "epoch": 1.9552715654952078,
+      "grad_norm": 4.860888016575171,
+      "learning_rate": 3.279228674928035e-06,
+      "loss": 0.6261,
+      "step": 612
+    },
+    {
+      "epoch": 1.9584664536741214,
+      "grad_norm": 2.8908433595575413,
+      "learning_rate": 3.261786864887117e-06,
+      "loss": 0.8123,
+      "step": 613
+    },
+    {
+      "epoch": 1.961661341853035,
+      "grad_norm": 3.3544693292462626,
+      "learning_rate": 3.244369081254585e-06,
+      "loss": 0.7071,
+      "step": 614
+    },
+    {
+      "epoch": 1.9648562300319488,
+      "grad_norm": 3.3673018356593745,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.6911,
+      "step": 615
+    },
+    {
+      "epoch": 1.9680511182108626,
+      "grad_norm": 3.1949674423232812,
+      "learning_rate": 3.209606555906788e-06,
+      "loss": 0.6891,
+      "step": 616
+    },
+    {
+      "epoch": 1.9712460063897763,
+      "grad_norm": 3.025487790758323,
+      "learning_rate": 3.192262294695679e-06,
+      "loss": 0.7605,
+      "step": 617
+    },
+    {
+      "epoch": 1.9744408945686902,
+      "grad_norm": 2.8013531181627984,
+      "learning_rate": 3.174943020894618e-06,
+      "loss": 0.5644,
+      "step": 618
+    },
+    {
+      "epoch": 1.9776357827476039,
+      "grad_norm": 3.202849902212168,
+      "learning_rate": 3.1576489738988457e-06,
+      "loss": 0.6496,
+      "step": 619
+    },
+    {
+      "epoch": 1.9808306709265175,
+      "grad_norm": 2.9966556403176905,
+      "learning_rate": 3.140380392754901e-06,
+      "loss": 0.8189,
+      "step": 620
+    },
+    {
+      "epoch": 1.9840255591054312,
+      "grad_norm": 3.387988163170543,
+      "learning_rate": 3.12313751615732e-06,
+      "loss": 0.6582,
+      "step": 621
+    },
+    {
+      "epoch": 1.9872204472843449,
+      "grad_norm": 3.060388847243892,
+      "learning_rate": 3.1059205824453446e-06,
+      "loss": 0.6042,
+      "step": 622
+    },
+    {
+      "epoch": 1.9904153354632588,
+      "grad_norm": 3.594240280376376,
+      "learning_rate": 3.0887298295996183e-06,
+      "loss": 0.7727,
+      "step": 623
+    },
+    {
+      "epoch": 1.9936102236421727,
+      "grad_norm": 3.306466864923194,
+      "learning_rate": 3.0715654952388957e-06,
+      "loss": 0.7377,
+      "step": 624
+    },
+    {
+      "epoch": 1.9968051118210863,
+      "grad_norm": 3.0502297597158723,
+      "learning_rate": 3.054427816616773e-06,
+      "loss": 0.7149,
+      "step": 625
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.0960499889219397,
+      "learning_rate": 3.0373170306183885e-06,
+      "loss": 0.5389,
+      "step": 626
+    },
+    {
+      "epoch": 2.0031948881789137,
+      "grad_norm": 3.983739654854693,
+      "learning_rate": 3.020233373757162e-06,
+      "loss": 0.3061,
+      "step": 627
+    },
+    {
+      "epoch": 2.0063897763578273,
+      "grad_norm": 3.285621566339913,
+      "learning_rate": 3.0031770821715233e-06,
+      "loss": 0.3634,
+      "step": 628
+    },
+    {
+      "epoch": 2.009584664536741,
+      "grad_norm": 3.3676704437636666,
+      "learning_rate": 2.9861483916216404e-06,
+      "loss": 0.3755,
+      "step": 629
+    },
+    {
+      "epoch": 2.012779552715655,
+      "grad_norm": 3.8878545954984483,
+      "learning_rate": 2.969147537486175e-06,
+      "loss": 0.3514,
+      "step": 630
+    },
+    {
+      "epoch": 2.015974440894569,
+      "grad_norm": 3.445732834591301,
+      "learning_rate": 2.952174754759012e-06,
+      "loss": 0.3014,
+      "step": 631
+    },
+    {
+      "epoch": 2.0191693290734825,
+      "grad_norm": 3.3721933738636207,
+      "learning_rate": 2.935230278046025e-06,
+      "loss": 0.27,
+      "step": 632
+    },
+    {
+      "epoch": 2.022364217252396,
+      "grad_norm": 3.1413868242299468,
+      "learning_rate": 2.9183143415618297e-06,
+      "loss": 0.4098,
+      "step": 633
+    },
+    {
+      "epoch": 2.02555910543131,
+      "grad_norm": 2.891829029150921,
+      "learning_rate": 2.9014271791265403e-06,
+      "loss": 0.4025,
+      "step": 634
+    },
+    {
+      "epoch": 2.0287539936102235,
+      "grad_norm": 3.026180083127087,
+      "learning_rate": 2.8845690241625437e-06,
+      "loss": 0.2911,
+      "step": 635
+    },
+    {
+      "epoch": 2.0319488817891376,
+      "grad_norm": 4.498220047068664,
+      "learning_rate": 2.867740109691277e-06,
+      "loss": 0.4535,
+      "step": 636
+    },
+    {
+      "epoch": 2.0351437699680512,
+      "grad_norm": 3.7345028381097474,
+      "learning_rate": 2.850940668329996e-06,
+      "loss": 0.351,
+      "step": 637
+    },
+    {
+      "epoch": 2.038338658146965,
+      "grad_norm": 3.2970944779006244,
+      "learning_rate": 2.8341709322885624e-06,
+      "loss": 0.3557,
+      "step": 638
+    },
+    {
+      "epoch": 2.0415335463258786,
+      "grad_norm": 4.895661775767794,
+      "learning_rate": 2.817431133366246e-06,
+      "loss": 0.4169,
+      "step": 639
+    },
+    {
+      "epoch": 2.0447284345047922,
+      "grad_norm": 4.021277358427465,
+      "learning_rate": 2.800721502948506e-06,
+      "loss": 0.3093,
+      "step": 640
+    },
+    {
+      "epoch": 2.047923322683706,
+      "grad_norm": 3.7231410410341157,
+      "learning_rate": 2.7840422720037943e-06,
+      "loss": 0.4976,
+      "step": 641
+    },
+    {
+      "epoch": 2.0511182108626196,
+      "grad_norm": 4.31493928366863,
+      "learning_rate": 2.767393671080376e-06,
+      "loss": 0.3381,
+      "step": 642
+    },
+    {
+      "epoch": 2.0543130990415337,
+      "grad_norm": 3.647720468261154,
+      "learning_rate": 2.7507759303031257e-06,
+      "loss": 0.3512,
+      "step": 643
+    },
+    {
+      "epoch": 2.0575079872204474,
+      "grad_norm": 3.330405855556239,
+      "learning_rate": 2.7341892793703594e-06,
+      "loss": 0.3616,
+      "step": 644
+    },
+    {
+      "epoch": 2.060702875399361,
+      "grad_norm": 3.6914736728228816,
+      "learning_rate": 2.7176339475506515e-06,
+      "loss": 0.3158,
+      "step": 645
+    },
+    {
+      "epoch": 2.0638977635782747,
+      "grad_norm": 3.3839418187871857,
+      "learning_rate": 2.7011101636796677e-06,
+      "loss": 0.2544,
+      "step": 646
+    },
+    {
+      "epoch": 2.0670926517571884,
+      "grad_norm": 3.117493894326818,
+      "learning_rate": 2.6846181561570085e-06,
+      "loss": 0.4699,
+      "step": 647
+    },
+    {
+      "epoch": 2.070287539936102,
+      "grad_norm": 2.5802684920101115,
+      "learning_rate": 2.668158152943039e-06,
+      "loss": 0.3334,
+      "step": 648
+    },
+    {
+      "epoch": 2.073482428115016,
+      "grad_norm": 2.997239923268133,
+      "learning_rate": 2.651730381555754e-06,
+      "loss": 0.3456,
+      "step": 649
+    },
+    {
+      "epoch": 2.07667731629393,
+      "grad_norm": 2.607698914364931,
+      "learning_rate": 2.635335069067617e-06,
+      "loss": 0.4099,
+      "step": 650
+    },
+    {
+      "epoch": 2.0798722044728435,
+      "grad_norm": 2.8928367127310657,
+      "learning_rate": 2.618972442102432e-06,
+      "loss": 0.4367,
+      "step": 651
+    },
+    {
+      "epoch": 2.083067092651757,
+      "grad_norm": 2.9734719232854974,
+      "learning_rate": 2.602642726832212e-06,
+      "loss": 0.3669,
+      "step": 652
+    },
+    {
+      "epoch": 2.086261980830671,
+      "grad_norm": 2.9938708892748203,
+      "learning_rate": 2.5863461489740403e-06,
+      "loss": 0.2868,
+      "step": 653
+    },
+    {
+      "epoch": 2.0894568690095845,
+      "grad_norm": 2.696302658673934,
+      "learning_rate": 2.57008293378697e-06,
+      "loss": 0.268,
+      "step": 654
+    },
+    {
+      "epoch": 2.0926517571884986,
+      "grad_norm": 3.2368753883570345,
+      "learning_rate": 2.553853306068888e-06,
+      "loss": 0.3144,
+      "step": 655
+    },
+    {
+      "epoch": 2.0958466453674123,
+      "grad_norm": 2.772114901987638,
+      "learning_rate": 2.5376574901534303e-06,
+      "loss": 0.3106,
+      "step": 656
+    },
+    {
+      "epoch": 2.099041533546326,
+      "grad_norm": 3.576165334680579,
+      "learning_rate": 2.5214957099068613e-06,
+      "loss": 0.3335,
+      "step": 657
+    },
+    {
+      "epoch": 2.1022364217252396,
+      "grad_norm": 2.8652986441263515,
+      "learning_rate": 2.5053681887249916e-06,
+      "loss": 0.2772,
+      "step": 658
+    },
+    {
+      "epoch": 2.1054313099041533,
+      "grad_norm": 2.8692228324598434,
+      "learning_rate": 2.4892751495300893e-06,
+      "loss": 0.3026,
+      "step": 659
+    },
+    {
+      "epoch": 2.108626198083067,
+      "grad_norm": 3.188288989732516,
+      "learning_rate": 2.4732168147677927e-06,
+      "loss": 0.3597,
+      "step": 660
+    },
+    {
+      "epoch": 2.1118210862619806,
+      "grad_norm": 2.8663933959362593,
+      "learning_rate": 2.4571934064040364e-06,
+      "loss": 0.2612,
+      "step": 661
+    },
+    {
+      "epoch": 2.1150159744408947,
+      "grad_norm": 3.4943865909460006,
+      "learning_rate": 2.4412051459219945e-06,
+      "loss": 0.3404,
+      "step": 662
+    },
+    {
+      "epoch": 2.1182108626198084,
+      "grad_norm": 3.178732383977166,
+      "learning_rate": 2.425252254319002e-06,
+      "loss": 0.3067,
+      "step": 663
+    },
+    {
+      "epoch": 2.121405750798722,
+      "grad_norm": 2.9321622441051267,
+      "learning_rate": 2.4093349521035105e-06,
+      "loss": 0.3245,
+      "step": 664
+    },
+    {
+      "epoch": 2.1246006389776357,
+      "grad_norm": 3.6061865707038896,
+      "learning_rate": 2.3934534592920416e-06,
+      "loss": 0.2906,
+      "step": 665
+    },
+    {
+      "epoch": 2.1277955271565494,
+      "grad_norm": 2.7151249200064833,
+      "learning_rate": 2.3776079954061385e-06,
+      "loss": 0.2442,
+      "step": 666
+    },
+    {
+      "epoch": 2.130990415335463,
+      "grad_norm": 3.2565886564034274,
+      "learning_rate": 2.3617987794693358e-06,
+      "loss": 0.3211,
+      "step": 667
+    },
+    {
+      "epoch": 2.134185303514377,
+      "grad_norm": 3.2348685094289644,
+      "learning_rate": 2.3460260300041355e-06,
+      "loss": 0.3291,
+      "step": 668
+    },
+    {
+      "epoch": 2.137380191693291,
+      "grad_norm": 2.558092203651786,
+      "learning_rate": 2.3302899650289773e-06,
+      "loss": 0.2296,
+      "step": 669
+    },
+    {
+      "epoch": 2.1405750798722045,
+      "grad_norm": 2.5276335316233722,
+      "learning_rate": 2.314590802055232e-06,
+      "loss": 0.2855,
+      "step": 670
+    },
+    {
+      "epoch": 2.143769968051118,
+      "grad_norm": 3.059331346532597,
+      "learning_rate": 2.2989287580841985e-06,
+      "loss": 0.2427,
+      "step": 671
+    },
+    {
+      "epoch": 2.146964856230032,
+      "grad_norm": 3.193721291328615,
+      "learning_rate": 2.2833040496040925e-06,
+      "loss": 0.2344,
+      "step": 672
+    },
+    {
+      "epoch": 2.1501597444089455,
+      "grad_norm": 4.163285212887703,
+      "learning_rate": 2.267716892587062e-06,
+      "loss": 0.3123,
+      "step": 673
+    },
+    {
+      "epoch": 2.1533546325878596,
+      "grad_norm": 3.3969901348176226,
+      "learning_rate": 2.252167502486205e-06,
+      "loss": 0.4806,
+      "step": 674
+    },
+    {
+      "epoch": 2.1565495207667733,
+      "grad_norm": 2.839343538628818,
+      "learning_rate": 2.2366560942325833e-06,
+      "loss": 0.3123,
+      "step": 675
+    },
+    {
+      "epoch": 2.159744408945687,
+      "grad_norm": 2.8099629116074447,
+      "learning_rate": 2.2211828822322547e-06,
+      "loss": 0.3208,
+      "step": 676
+    },
+    {
+      "epoch": 2.1629392971246006,
+      "grad_norm": 2.7820893478354325,
+      "learning_rate": 2.205748080363316e-06,
+      "loss": 0.2558,
+      "step": 677
+    },
+    {
+      "epoch": 2.1661341853035143,
+      "grad_norm": 3.6656201908136614,
+      "learning_rate": 2.190351901972935e-06,
+      "loss": 0.3526,
+      "step": 678
+    },
+    {
+      "epoch": 2.169329073482428,
+      "grad_norm": 3.128153155659913,
+      "learning_rate": 2.1749945598744076e-06,
+      "loss": 0.3192,
+      "step": 679
+    },
+    {
+      "epoch": 2.1725239616613417,
+      "grad_norm": 3.2972703109935986,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.4369,
+      "step": 680
+    },
+    {
+      "epoch": 2.1757188498402558,
+      "grad_norm": 2.424788992475203,
+      "learning_rate": 2.144397233119112e-06,
+      "loss": 0.3038,
+      "step": 681
+    },
+    {
+      "epoch": 2.1789137380191694,
+      "grad_norm": 3.224412274606723,
+      "learning_rate": 2.1291576713931382e-06,
+      "loss": 0.3845,
+      "step": 682
+    },
+    {
+      "epoch": 2.182108626198083,
+      "grad_norm": 2.8590636348392096,
+      "learning_rate": 2.1139577918147715e-06,
+      "loss": 0.3982,
+      "step": 683
+    },
+    {
+      "epoch": 2.1853035143769968,
+      "grad_norm": 2.9859896483829558,
+      "learning_rate": 2.0987978044839707e-06,
+      "loss": 0.4328,
+      "step": 684
+    },
+    {
+      "epoch": 2.1884984025559104,
+      "grad_norm": 3.8107950411886518,
+      "learning_rate": 2.0836779189492925e-06,
+      "loss": 0.3786,
+      "step": 685
+    },
+    {
+      "epoch": 2.191693290734824,
+      "grad_norm": 3.1009875807967924,
+      "learning_rate": 2.068598344204981e-06,
+      "loss": 0.5021,
+      "step": 686
+    },
+    {
+      "epoch": 2.194888178913738,
+      "grad_norm": 2.3440228686710065,
+      "learning_rate": 2.053559288688086e-06,
+      "loss": 0.2171,
+      "step": 687
+    },
+    {
+      "epoch": 2.198083067092652,
+      "grad_norm": 3.5294078738520125,
+      "learning_rate": 2.0385609602755878e-06,
+      "loss": 0.257,
+      "step": 688
+    },
+    {
+      "epoch": 2.2012779552715656,
+      "grad_norm": 3.0790750421176205,
+      "learning_rate": 2.02360356628151e-06,
+      "loss": 0.3182,
+      "step": 689
+    },
+    {
+      "epoch": 2.2044728434504792,
+      "grad_norm": 5.154029864035265,
+      "learning_rate": 2.0086873134540626e-06,
+      "loss": 0.2653,
+      "step": 690
+    },
+    {
+      "epoch": 2.207667731629393,
+      "grad_norm": 2.8543026187870004,
+      "learning_rate": 1.9938124079727874e-06,
+      "loss": 0.3174,
+      "step": 691
+    },
+    {
+      "epoch": 2.2108626198083066,
+      "grad_norm": 2.6351708701844014,
+      "learning_rate": 1.9789790554456977e-06,
+      "loss": 0.2985,
+      "step": 692
+    },
+    {
+      "epoch": 2.2140575079872207,
+      "grad_norm": 2.5496074335825907,
+      "learning_rate": 1.9641874609064443e-06,
+      "loss": 0.2142,
+      "step": 693
+    },
+    {
+      "epoch": 2.2172523961661343,
+      "grad_norm": 3.03610814609126,
+      "learning_rate": 1.9494378288114816e-06,
+      "loss": 0.3684,
+      "step": 694
+    },
+    {
+      "epoch": 2.220447284345048,
+      "grad_norm": 2.4059048003356214,
+      "learning_rate": 1.9347303630372373e-06,
+      "loss": 0.2973,
+      "step": 695
+    },
+    {
+      "epoch": 2.2236421725239617,
+      "grad_norm": 2.737026725515166,
+      "learning_rate": 1.9200652668772924e-06,
+      "loss": 0.3047,
+      "step": 696
+    },
+    {
+      "epoch": 2.2268370607028753,
+      "grad_norm": 2.4710903388986596,
+      "learning_rate": 1.9054427430395828e-06,
+      "loss": 0.2045,
+      "step": 697
+    },
+    {
+      "epoch": 2.230031948881789,
+      "grad_norm": 2.6582459879412315,
+      "learning_rate": 1.890862993643583e-06,
+      "loss": 0.2705,
+      "step": 698
+    },
+    {
+      "epoch": 2.2332268370607027,
+      "grad_norm": 3.036328419292376,
+      "learning_rate": 1.8763262202175204e-06,
+      "loss": 0.2761,
+      "step": 699
+    },
+    {
+      "epoch": 2.236421725239617,
+      "grad_norm": 4.148400766996593,
+      "learning_rate": 1.8618326236955908e-06,
+      "loss": 0.2652,
+      "step": 700
+    },
+    {
+      "epoch": 2.2396166134185305,
+      "grad_norm": 3.5888332943960823,
+      "learning_rate": 1.8473824044151762e-06,
+      "loss": 0.4339,
+      "step": 701
+    },
+    {
+      "epoch": 2.242811501597444,
+      "grad_norm": 2.6608565764460805,
+      "learning_rate": 1.8329757621140748e-06,
+      "loss": 0.2986,
+      "step": 702
+    },
+    {
+      "epoch": 2.246006389776358,
+      "grad_norm": 2.868795114522909,
+      "learning_rate": 1.81861289592775e-06,
+      "loss": 0.3452,
+      "step": 703
+    },
+    {
+      "epoch": 2.2492012779552715,
+      "grad_norm": 3.4182446194718388,
+      "learning_rate": 1.8042940043865658e-06,
+      "loss": 0.3306,
+      "step": 704
+    },
+    {
+      "epoch": 2.252396166134185,
+      "grad_norm": 2.675796053709656,
+      "learning_rate": 1.7900192854130465e-06,
+      "loss": 0.3718,
+      "step": 705
+    },
+    {
+      "epoch": 2.255591054313099,
+      "grad_norm": 2.8152112447207203,
+      "learning_rate": 1.7757889363191484e-06,
+      "loss": 0.3105,
+      "step": 706
+    },
+    {
+      "epoch": 2.258785942492013,
+      "grad_norm": 2.517527215822727,
+      "learning_rate": 1.7616031538035189e-06,
+      "loss": 0.1474,
+      "step": 707
+    },
+    {
+      "epoch": 2.2619808306709266,
+      "grad_norm": 3.5349716425124016,
+      "learning_rate": 1.7474621339487925e-06,
+      "loss": 0.3869,
+      "step": 708
+    },
+    {
+      "epoch": 2.2651757188498403,
+      "grad_norm": 2.9091370924854902,
+      "learning_rate": 1.7333660722188667e-06,
+      "loss": 0.2907,
+      "step": 709
+    },
+    {
+      "epoch": 2.268370607028754,
+      "grad_norm": 2.938936979003365,
+      "learning_rate": 1.7193151634562071e-06,
+      "loss": 0.3422,
+      "step": 710
+    },
+    {
+      "epoch": 2.2715654952076676,
+      "grad_norm": 3.2492191599631783,
+      "learning_rate": 1.7053096018791588e-06,
+      "loss": 0.3079,
+      "step": 711
+    },
+    {
+      "epoch": 2.2747603833865817,
+      "grad_norm": 2.7744197458417332,
+      "learning_rate": 1.691349581079249e-06,
+      "loss": 0.2767,
+      "step": 712
+    },
+    {
+      "epoch": 2.2779552715654954,
+      "grad_norm": 2.401629033687684,
+      "learning_rate": 1.6774352940185269e-06,
+      "loss": 0.3173,
+      "step": 713
+    },
+    {
+      "epoch": 2.281150159744409,
+      "grad_norm": 3.204018555728656,
+      "learning_rate": 1.663566933026879e-06,
+      "loss": 0.3041,
+      "step": 714
+    },
+    {
+      "epoch": 2.2843450479233227,
+      "grad_norm": 2.6659318029935672,
+      "learning_rate": 1.6497446897993885e-06,
+      "loss": 0.3293,
+      "step": 715
+    },
+    {
+      "epoch": 2.2875399361022364,
+      "grad_norm": 2.3994693364686066,
+      "learning_rate": 1.6359687553936714e-06,
+      "loss": 0.3874,
+      "step": 716
+    },
+    {
+      "epoch": 2.29073482428115,
+      "grad_norm": 3.641566462537511,
+      "learning_rate": 1.6222393202272414e-06,
+      "loss": 0.4569,
+      "step": 717
+    },
+    {
+      "epoch": 2.2939297124600637,
+      "grad_norm": 2.566776293738956,
+      "learning_rate": 1.6085565740748825e-06,
+      "loss": 0.3578,
+      "step": 718
+    },
+    {
+      "epoch": 2.297124600638978,
+      "grad_norm": 2.8020227412037784,
+      "learning_rate": 1.5949207060660138e-06,
+      "loss": 0.2513,
+      "step": 719
+    },
+    {
+      "epoch": 2.3003194888178915,
+      "grad_norm": 2.6446468775339875,
+      "learning_rate": 1.581331904682089e-06,
+      "loss": 0.2321,
+      "step": 720
+    },
+    {
+      "epoch": 2.303514376996805,
+      "grad_norm": 2.6700679646491765,
+      "learning_rate": 1.5677903577539806e-06,
+      "loss": 0.2647,
+      "step": 721
+    },
+    {
+      "epoch": 2.306709265175719,
+      "grad_norm": 2.716028000148245,
+      "learning_rate": 1.5542962524593869e-06,
+      "loss": 0.3478,
+      "step": 722
+    },
+    {
+      "epoch": 2.3099041533546325,
+      "grad_norm": 3.1191825295408946,
+      "learning_rate": 1.54084977532025e-06,
+      "loss": 0.2332,
+      "step": 723
+    },
+    {
+      "epoch": 2.313099041533546,
+      "grad_norm": 3.122396376234942,
+      "learning_rate": 1.5274511122001684e-06,
+      "loss": 0.383,
+      "step": 724
+    },
+    {
+      "epoch": 2.31629392971246,
+      "grad_norm": 2.9827586981366765,
+      "learning_rate": 1.5141004483018323e-06,
+      "loss": 0.3597,
+      "step": 725
+    },
+    {
+      "epoch": 2.319488817891374,
+      "grad_norm": 3.4251031092935618,
+      "learning_rate": 1.5007979681644696e-06,
+      "loss": 0.1876,
+      "step": 726
+    },
+    {
+      "epoch": 2.3226837060702876,
+      "grad_norm": 3.2332034094836954,
+      "learning_rate": 1.4875438556612836e-06,
+      "loss": 0.4144,
+      "step": 727
+    },
+    {
+      "epoch": 2.3258785942492013,
+      "grad_norm": 3.571622683368131,
+      "learning_rate": 1.474338293996917e-06,
+      "loss": 0.3402,
+      "step": 728
+    },
+    {
+      "epoch": 2.329073482428115,
+      "grad_norm": 2.6498229979086796,
+      "learning_rate": 1.4611814657049257e-06,
+      "loss": 0.3,
+      "step": 729
+    },
+    {
+      "epoch": 2.3322683706070286,
+      "grad_norm": 2.9332089675275763,
+      "learning_rate": 1.4480735526452427e-06,
+      "loss": 0.2661,
+      "step": 730
+    },
+    {
+      "epoch": 2.3354632587859427,
+      "grad_norm": 3.64974422306179,
+      "learning_rate": 1.4350147360016743e-06,
+      "loss": 0.3723,
+      "step": 731
+    },
+    {
+      "epoch": 2.3386581469648564,
+      "grad_norm": 3.2718140557881683,
+      "learning_rate": 1.4220051962793952e-06,
+      "loss": 0.341,
+      "step": 732
+    },
+    {
+      "epoch": 2.34185303514377,
+      "grad_norm": 4.027162305240513,
+      "learning_rate": 1.4090451133024473e-06,
+      "loss": 0.3616,
+      "step": 733
+    },
+    {
+      "epoch": 2.3450479233226837,
+      "grad_norm": 2.9360663676398184,
+      "learning_rate": 1.3961346662112585e-06,
+      "loss": 0.3221,
+      "step": 734
+    },
+    {
+      "epoch": 2.3482428115015974,
+      "grad_norm": 2.56996769881666,
+      "learning_rate": 1.3832740334601692e-06,
+      "loss": 0.3172,
+      "step": 735
+    },
+    {
+      "epoch": 2.351437699680511,
+      "grad_norm": 3.00634707389068,
+      "learning_rate": 1.3704633928149575e-06,
+      "loss": 0.3467,
+      "step": 736
+    },
+    {
+      "epoch": 2.3546325878594248,
+      "grad_norm": 2.957941155180034,
+      "learning_rate": 1.3577029213503911e-06,
+      "loss": 0.3569,
+      "step": 737
+    },
+    {
+      "epoch": 2.357827476038339,
+      "grad_norm": 3.340715043128515,
+      "learning_rate": 1.3449927954477732e-06,
+      "loss": 0.3943,
+      "step": 738
+    },
+    {
+      "epoch": 2.3610223642172525,
+      "grad_norm": 2.7404476629806562,
+      "learning_rate": 1.3323331907925046e-06,
+      "loss": 0.3187,
+      "step": 739
+    },
+    {
+      "epoch": 2.364217252396166,
+      "grad_norm": 2.5246156692846875,
+      "learning_rate": 1.319724282371664e-06,
+      "loss": 0.2475,
+      "step": 740
+    },
+    {
+      "epoch": 2.36741214057508,
+      "grad_norm": 2.873519654865862,
+      "learning_rate": 1.307166244471576e-06,
+      "loss": 0.3298,
+      "step": 741
+    },
+    {
+      "epoch": 2.3706070287539935,
+      "grad_norm": 2.390732500267084,
+      "learning_rate": 1.2946592506754097e-06,
+      "loss": 0.3296,
+      "step": 742
+    },
+    {
+      "epoch": 2.373801916932907,
+      "grad_norm": 3.1818240368904855,
+      "learning_rate": 1.282203473860783e-06,
+      "loss": 0.3843,
+      "step": 743
+    },
+    {
+      "epoch": 2.376996805111821,
+      "grad_norm": 2.4517581011937684,
+      "learning_rate": 1.2697990861973635e-06,
+      "loss": 0.2315,
+      "step": 744
+    },
+    {
+      "epoch": 2.380191693290735,
+      "grad_norm": 2.573546354242588,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.2236,
+      "step": 745
+    },
+    {
+      "epoch": 2.3833865814696487,
+      "grad_norm": 2.8468269899380614,
+      "learning_rate": 1.2451451634488264e-06,
+      "loss": 0.2713,
+      "step": 746
+    },
+    {
+      "epoch": 2.3865814696485623,
+      "grad_norm": 2.588569151009441,
+      "learning_rate": 1.2328959691419517e-06,
+      "loss": 0.3862,
+      "step": 747
+    },
+    {
+      "epoch": 2.389776357827476,
+      "grad_norm": 2.839149450846035,
+      "learning_rate": 1.2206988455380558e-06,
+      "loss": 0.4483,
+      "step": 748
+    },
+    {
+      "epoch": 2.3929712460063897,
+      "grad_norm": 2.9746150005627396,
+      "learning_rate": 1.2085539612315844e-06,
+      "loss": 0.4025,
+      "step": 749
+    },
+    {
+      "epoch": 2.3961661341853033,
+      "grad_norm": 2.623868573575473,
+      "learning_rate": 1.1964614840949002e-06,
+      "loss": 0.3016,
+      "step": 750
+    },
+    {
+      "epoch": 2.3993610223642174,
+      "grad_norm": 3.031338624058212,
+      "learning_rate": 1.1844215812759708e-06,
+      "loss": 0.2565,
+      "step": 751
+    },
+    {
+      "epoch": 2.402555910543131,
+      "grad_norm": 2.5071778757846936,
+      "learning_rate": 1.1724344191960591e-06,
+      "loss": 0.4643,
+      "step": 752
+    },
+    {
+      "epoch": 2.405750798722045,
+      "grad_norm": 2.8317059479994886,
+      "learning_rate": 1.1605001635474183e-06,
+      "loss": 0.2481,
+      "step": 753
+    },
+    {
+      "epoch": 2.4089456869009584,
+      "grad_norm": 3.102116970392882,
+      "learning_rate": 1.1486189792910024e-06,
+      "loss": 0.3586,
+      "step": 754
+    },
+    {
+      "epoch": 2.412140575079872,
+      "grad_norm": 2.6451678804583847,
+      "learning_rate": 1.1367910306541918e-06,
+      "loss": 0.2756,
+      "step": 755
+    },
+    {
+      "epoch": 2.415335463258786,
+      "grad_norm": 4.086811018849218,
+      "learning_rate": 1.1250164811285148e-06,
+      "loss": 0.2683,
+      "step": 756
+    },
+    {
+      "epoch": 2.4185303514377,
+      "grad_norm": 2.600333389040635,
+      "learning_rate": 1.1132954934673911e-06,
+      "loss": 0.2296,
+      "step": 757
+    },
+    {
+      "epoch": 2.4217252396166136,
+      "grad_norm": 3.967573333157087,
+      "learning_rate": 1.1016282296838887e-06,
+      "loss": 0.36,
+      "step": 758
+    },
+    {
+      "epoch": 2.4249201277955272,
+      "grad_norm": 2.233497426709634,
+      "learning_rate": 1.090014851048473e-06,
+      "loss": 0.2362,
+      "step": 759
+    },
+    {
+      "epoch": 2.428115015974441,
+      "grad_norm": 2.5925854068065415,
+      "learning_rate": 1.078455518086784e-06,
+      "loss": 0.2482,
+      "step": 760
+    },
+    {
+      "epoch": 2.4313099041533546,
+      "grad_norm": 4.587064754811665,
+      "learning_rate": 1.0669503905774198e-06,
+      "loss": 0.2976,
+      "step": 761
+    },
+    {
+      "epoch": 2.4345047923322682,
+      "grad_norm": 3.416262667137309,
+      "learning_rate": 1.055499627549722e-06,
+      "loss": 0.3512,
+      "step": 762
+    },
+    {
+      "epoch": 2.437699680511182,
+      "grad_norm": 3.426620474014868,
+      "learning_rate": 1.0441033872815804e-06,
+      "loss": 0.2857,
+      "step": 763
+    },
+    {
+      "epoch": 2.440894568690096,
+      "grad_norm": 4.009236379600599,
+      "learning_rate": 1.0327618272972484e-06,
+      "loss": 0.321,
+      "step": 764
+    },
+    {
+      "epoch": 2.4440894568690097,
+      "grad_norm": 3.572224194215062,
+      "learning_rate": 1.0214751043651582e-06,
+      "loss": 0.3032,
+      "step": 765
+    },
+    {
+      "epoch": 2.4472843450479234,
+      "grad_norm": 3.124167617270443,
+      "learning_rate": 1.010243374495763e-06,
+      "loss": 0.3955,
+      "step": 766
+    },
+    {
+      "epoch": 2.450479233226837,
+      "grad_norm": 2.6229893932975714,
+      "learning_rate": 9.990667929393715e-07,
+      "loss": 0.3052,
+      "step": 767
+    },
+    {
+      "epoch": 2.4536741214057507,
+      "grad_norm": 2.8816075910274788,
+      "learning_rate": 9.879455141840067e-07,
+      "loss": 0.44,
+      "step": 768
+    },
+    {
+      "epoch": 2.4568690095846644,
+      "grad_norm": 4.112012621253609,
+      "learning_rate": 9.768796919532742e-07,
+      "loss": 0.2652,
+      "step": 769
+    },
+    {
+      "epoch": 2.460063897763578,
+      "grad_norm": 2.648882925688484,
+      "learning_rate": 9.658694792042284e-07,
+      "loss": 0.4499,
+      "step": 770
+    },
+    {
+      "epoch": 2.463258785942492,
+      "grad_norm": 3.1547924763584962,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.3785,
+      "step": 771
+    },
+    {
+      "epoch": 2.466453674121406,
+      "grad_norm": 2.9952168763350984,
+      "learning_rate": 9.440164901340127e-07,
+      "loss": 0.3601,
+      "step": 772
+    },
+    {
+      "epoch": 2.4696485623003195,
+      "grad_norm": 2.8146008583165893,
+      "learning_rate": 9.331740158752495e-07,
+      "loss": 0.2369,
+      "step": 773
+    },
+    {
+      "epoch": 2.472843450479233,
+      "grad_norm": 2.277449084789617,
+      "learning_rate": 9.223877552188065e-07,
+      "loss": 0.2875,
+      "step": 774
+    },
+    {
+      "epoch": 2.476038338658147,
+      "grad_norm": 2.359193486193659,
+      "learning_rate": 9.116578572575091e-07,
+      "loss": 0.2897,
+      "step": 775
+    },
+    {
+      "epoch": 2.479233226837061,
+      "grad_norm": 3.060706154822991,
+      "learning_rate": 9.009844703051063e-07,
+      "loss": 0.328,
+      "step": 776
+    },
+    {
+      "epoch": 2.4824281150159746,
+      "grad_norm": 3.4076538680190946,
+      "learning_rate": 8.903677418942292e-07,
+      "loss": 0.3422,
+      "step": 777
+    },
+    {
+      "epoch": 2.4856230031948883,
+      "grad_norm": 2.5012479003656667,
+      "learning_rate": 8.79807818774343e-07,
+      "loss": 0.3489,
+      "step": 778
+    },
+    {
+      "epoch": 2.488817891373802,
+      "grad_norm": 2.6471882115948486,
+      "learning_rate": 8.693048469097293e-07,
+      "loss": 0.2347,
+      "step": 779
+    },
+    {
+      "epoch": 2.4920127795527156,
+      "grad_norm": 2.563868274335942,
+      "learning_rate": 8.58858971477457e-07,
+      "loss": 0.26,
+      "step": 780
+    },
+    {
+      "epoch": 2.4952076677316293,
+      "grad_norm": 3.4646909340715455,
+      "learning_rate": 8.484703368653812e-07,
+      "loss": 0.3294,
+      "step": 781
+    },
+    {
+      "epoch": 2.498402555910543,
+      "grad_norm": 2.679833172377346,
+      "learning_rate": 8.381390866701517e-07,
+      "loss": 0.2815,
+      "step": 782
+    },
+    {
+      "epoch": 2.501597444089457,
+      "grad_norm": 2.9115947719650395,
+      "learning_rate": 8.278653636952177e-07,
+      "loss": 0.3526,
+      "step": 783
+    },
+    {
+      "epoch": 2.5047923322683707,
+      "grad_norm": 2.9410341990050743,
+      "learning_rate": 8.176493099488664e-07,
+      "loss": 0.2946,
+      "step": 784
+    },
+    {
+      "epoch": 2.5079872204472844,
+      "grad_norm": 2.963777621766596,
+      "learning_rate": 8.074910666422475e-07,
+      "loss": 0.3603,
+      "step": 785
+    },
+    {
+      "epoch": 2.511182108626198,
+      "grad_norm": 2.5982819562264563,
+      "learning_rate": 7.973907741874287e-07,
+      "loss": 0.3086,
+      "step": 786
+    },
+    {
+      "epoch": 2.5143769968051117,
+      "grad_norm": 2.7398047705319963,
+      "learning_rate": 7.873485721954572e-07,
+      "loss": 0.2696,
+      "step": 787
+    },
+    {
+      "epoch": 2.5175718849840254,
+      "grad_norm": 3.2638294463866533,
+      "learning_rate": 7.773645994744222e-07,
+      "loss": 0.271,
+      "step": 788
+    },
+    {
+      "epoch": 2.520766773162939,
+      "grad_norm": 2.8576375719959373,
+      "learning_rate": 7.674389940275406e-07,
+      "loss": 0.3681,
+      "step": 789
+    },
+    {
+      "epoch": 2.523961661341853,
+      "grad_norm": 2.7374633501214967,
+      "learning_rate": 7.575718930512516e-07,
+      "loss": 0.1814,
+      "step": 790
+    },
+    {
+      "epoch": 2.527156549520767,
+      "grad_norm": 2.714207771128787,
+      "learning_rate": 7.47763432933315e-07,
+      "loss": 0.2809,
+      "step": 791
+    },
+    {
+      "epoch": 2.5303514376996805,
+      "grad_norm": 2.8312290802610156,
+      "learning_rate": 7.380137492509309e-07,
+      "loss": 0.2833,
+      "step": 792
+    },
+    {
+      "epoch": 2.533546325878594,
+      "grad_norm": 3.054754873253588,
+      "learning_rate": 7.283229767688627e-07,
+      "loss": 0.2806,
+      "step": 793
+    },
+    {
+      "epoch": 2.536741214057508,
+      "grad_norm": 2.828286671334927,
+      "learning_rate": 7.186912494375736e-07,
+      "loss": 0.3156,
+      "step": 794
+    },
+    {
+      "epoch": 2.539936102236422,
+      "grad_norm": 2.8829752863276488,
+      "learning_rate": 7.091187003913802e-07,
+      "loss": 0.3443,
+      "step": 795
+    },
+    {
+      "epoch": 2.543130990415335,
+      "grad_norm": 3.053676055897468,
+      "learning_rate": 6.996054619466053e-07,
+      "loss": 0.3243,
+      "step": 796
+    },
+    {
+      "epoch": 2.5463258785942493,
+      "grad_norm": 3.0660416023507326,
+      "learning_rate": 6.901516655997536e-07,
+      "loss": 0.3214,
+      "step": 797
+    },
+    {
+      "epoch": 2.549520766773163,
+      "grad_norm": 2.7066895892262783,
+      "learning_rate": 6.80757442025694e-07,
+      "loss": 0.1718,
+      "step": 798
+    },
+    {
+      "epoch": 2.5527156549520766,
+      "grad_norm": 2.603409237479928,
+      "learning_rate": 6.714229210758516e-07,
+      "loss": 0.2741,
+      "step": 799
+    },
+    {
+      "epoch": 2.5559105431309903,
+      "grad_norm": 3.0966996686574695,
+      "learning_rate": 6.621482317764105e-07,
+      "loss": 0.3319,
+      "step": 800
+    },
+    {
+      "epoch": 2.559105431309904,
+      "grad_norm": 2.6219749623887965,
+      "learning_rate": 6.529335023265387e-07,
+      "loss": 0.3564,
+      "step": 801
+    },
+    {
+      "epoch": 2.562300319488818,
+      "grad_norm": 3.086650960618863,
+      "learning_rate": 6.437788600966066e-07,
+      "loss": 0.3018,
+      "step": 802
+    },
+    {
+      "epoch": 2.5654952076677318,
+      "grad_norm": 3.1116723092796383,
+      "learning_rate": 6.346844316264312e-07,
+      "loss": 0.298,
+      "step": 803
+    },
+    {
+      "epoch": 2.5686900958466454,
+      "grad_norm": 2.411312374734287,
+      "learning_rate": 6.256503426235277e-07,
+      "loss": 0.244,
+      "step": 804
+    },
+    {
+      "epoch": 2.571884984025559,
+      "grad_norm": 3.6203829111687664,
+      "learning_rate": 6.166767179613691e-07,
+      "loss": 0.3591,
+      "step": 805
+    },
+    {
+      "epoch": 2.5750798722044728,
+      "grad_norm": 2.7876507873844614,
+      "learning_rate": 6.077636816776611e-07,
+      "loss": 0.2565,
+      "step": 806
+    },
+    {
+      "epoch": 2.5782747603833864,
+      "grad_norm": 3.015179861020195,
+      "learning_rate": 5.989113569726312e-07,
+      "loss": 0.3323,
+      "step": 807
+    },
+    {
+      "epoch": 2.5814696485623,
+      "grad_norm": 2.9444097234981395,
+      "learning_rate": 5.901198662073188e-07,
+      "loss": 0.2723,
+      "step": 808
+    },
+    {
+      "epoch": 2.584664536741214,
+      "grad_norm": 3.7185269730357087,
+      "learning_rate": 5.813893309018881e-07,
+      "loss": 0.3447,
+      "step": 809
+    },
+    {
+      "epoch": 2.587859424920128,
+      "grad_norm": 2.225048766195535,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.174,
+      "step": 810
+    },
+    {
+      "epoch": 2.5910543130990416,
+      "grad_norm": 3.44474391508872,
+      "learning_rate": 5.641116085368931e-07,
+      "loss": 0.2253,
+      "step": 811
+    },
+    {
+      "epoch": 2.594249201277955,
+      "grad_norm": 2.448094711116419,
+      "learning_rate": 5.555646602982207e-07,
+      "loss": 0.2955,
+      "step": 812
+    },
+    {
+      "epoch": 2.597444089456869,
+      "grad_norm": 2.563501210407334,
+      "learning_rate": 5.470791451579172e-07,
+      "loss": 0.3535,
+      "step": 813
+    },
+    {
+      "epoch": 2.600638977635783,
+      "grad_norm": 2.9291017278425455,
+      "learning_rate": 5.386551804068063e-07,
+      "loss": 0.3538,
+      "step": 814
+    },
+    {
+      "epoch": 2.6038338658146962,
+      "grad_norm": 2.7528795248116826,
+      "learning_rate": 5.302928824849335e-07,
+      "loss": 0.2973,
+      "step": 815
+    },
+    {
+      "epoch": 2.6070287539936103,
+      "grad_norm": 2.7382446313386315,
+      "learning_rate": 5.219923669799587e-07,
+      "loss": 0.2157,
+      "step": 816
+    },
+    {
+      "epoch": 2.610223642172524,
+      "grad_norm": 3.4302073328410256,
+      "learning_rate": 5.137537486255517e-07,
+      "loss": 0.314,
+      "step": 817
+    },
+    {
+      "epoch": 2.6134185303514377,
+      "grad_norm": 3.716094837100623,
+      "learning_rate": 5.055771412998122e-07,
+      "loss": 0.4665,
+      "step": 818
+    },
+    {
+      "epoch": 2.6166134185303513,
+      "grad_norm": 2.7131645523624366,
+      "learning_rate": 4.974626580236957e-07,
+      "loss": 0.2432,
+      "step": 819
+    },
+    {
+      "epoch": 2.619808306709265,
+      "grad_norm": 3.773747007444739,
+      "learning_rate": 4.894104109594466e-07,
+      "loss": 0.4126,
+      "step": 820
+    },
+    {
+      "epoch": 2.623003194888179,
+      "grad_norm": 3.480108229172227,
+      "learning_rate": 4.814205114090543e-07,
+      "loss": 0.2425,
+      "step": 821
+    },
+    {
+      "epoch": 2.626198083067093,
+      "grad_norm": 3.3210545461804823,
+      "learning_rate": 4.734930698127077e-07,
+      "loss": 0.2516,
+      "step": 822
+    },
+    {
+      "epoch": 2.6293929712460065,
+      "grad_norm": 2.509780782611665,
+      "learning_rate": 4.6562819574727304e-07,
+      "loss": 0.1837,
+      "step": 823
+    },
+    {
+      "epoch": 2.63258785942492,
+      "grad_norm": 2.850402113716751,
+      "learning_rate": 4.578259979247801e-07,
+      "loss": 0.3563,
+      "step": 824
+    },
+    {
+      "epoch": 2.635782747603834,
+      "grad_norm": 4.544846820980652,
+      "learning_rate": 4.500865841909169e-07,
+      "loss": 0.2435,
+      "step": 825
+    },
+    {
+      "epoch": 2.6389776357827475,
+      "grad_norm": 3.200856834109285,
+      "learning_rate": 4.4241006152353885e-07,
+      "loss": 0.4451,
+      "step": 826
+    },
+    {
+      "epoch": 2.642172523961661,
+      "grad_norm": 3.4524999136389174,
+      "learning_rate": 4.3479653603119287e-07,
+      "loss": 0.2656,
+      "step": 827
+    },
+    {
+      "epoch": 2.6453674121405752,
+      "grad_norm": 2.5305663904889357,
+      "learning_rate": 4.2724611295164755e-07,
+      "loss": 0.3413,
+      "step": 828
+    },
+    {
+      "epoch": 2.648562300319489,
+      "grad_norm": 3.819885095480714,
+      "learning_rate": 4.197588966504401e-07,
+      "loss": 0.295,
+      "step": 829
+    },
+    {
+      "epoch": 2.6517571884984026,
+      "grad_norm": 3.84940870071906,
+      "learning_rate": 4.123349906194357e-07,
+      "loss": 0.3689,
+      "step": 830
+    },
+    {
+      "epoch": 2.6549520766773163,
+      "grad_norm": 2.629296710525597,
+      "learning_rate": 4.0497449747539217e-07,
+      "loss": 0.4816,
+      "step": 831
+    },
+    {
+      "epoch": 2.65814696485623,
+      "grad_norm": 2.972520920912528,
+      "learning_rate": 3.9767751895854467e-07,
+      "loss": 0.2935,
+      "step": 832
+    },
+    {
+      "epoch": 2.661341853035144,
+      "grad_norm": 3.0988250894778586,
+      "learning_rate": 3.904441559312006e-07,
+      "loss": 0.3341,
+      "step": 833
+    },
+    {
+      "epoch": 2.6645367412140573,
+      "grad_norm": 3.7579842135072403,
+      "learning_rate": 3.8327450837634284e-07,
+      "loss": 0.3298,
+      "step": 834
+    },
+    {
+      "epoch": 2.6677316293929714,
+      "grad_norm": 2.6682110481776404,
+      "learning_rate": 3.7616867539624733e-07,
+      "loss": 0.2616,
+      "step": 835
+    },
+    {
+      "epoch": 2.670926517571885,
+      "grad_norm": 3.1415039719767535,
+      "learning_rate": 3.691267552111183e-07,
+      "loss": 0.3848,
+      "step": 836
+    },
+    {
+      "epoch": 2.6741214057507987,
+      "grad_norm": 2.7594774830021223,
+      "learning_rate": 3.621488451577221e-07,
+      "loss": 0.3824,
+      "step": 837
+    },
+    {
+      "epoch": 2.6773162939297124,
+      "grad_norm": 3.200662433005896,
+      "learning_rate": 3.552350416880507e-07,
+      "loss": 0.3202,
+      "step": 838
+    },
+    {
+      "epoch": 2.680511182108626,
+      "grad_norm": 2.6991157794601133,
+      "learning_rate": 3.483854403679832e-07,
+      "loss": 0.2799,
+      "step": 839
+    },
+    {
+      "epoch": 2.68370607028754,
+      "grad_norm": 3.1931633841331304,
+      "learning_rate": 3.416001358759635e-07,
+      "loss": 0.3718,
+      "step": 840
+    },
+    {
+      "epoch": 2.686900958466454,
+      "grad_norm": 2.6393238967880794,
+      "learning_rate": 3.3487922200169944e-07,
+      "loss": 0.2879,
+      "step": 841
+    },
+    {
+      "epoch": 2.6900958466453675,
+      "grad_norm": 2.8451904737691884,
+      "learning_rate": 3.2822279164485494e-07,
+      "loss": 0.2858,
+      "step": 842
+    },
+    {
+      "epoch": 2.693290734824281,
+      "grad_norm": 2.9497857503016798,
+      "learning_rate": 3.2163093681377765e-07,
+      "loss": 0.4216,
+      "step": 843
+    },
+    {
+      "epoch": 2.696485623003195,
+      "grad_norm": 4.216992352776779,
+      "learning_rate": 3.151037486242181e-07,
+      "loss": 0.2952,
+      "step": 844
+    },
+    {
+      "epoch": 2.6996805111821085,
+      "grad_norm": 2.9505970699645716,
+      "learning_rate": 3.08641317298074e-07,
+      "loss": 0.3443,
+      "step": 845
+    },
+    {
+      "epoch": 2.702875399361022,
+      "grad_norm": 2.3456532898336184,
+      "learning_rate": 3.022437321621452e-07,
+      "loss": 0.2235,
+      "step": 846
+    },
+    {
+      "epoch": 2.7060702875399363,
+      "grad_norm": 2.9087094972449017,
+      "learning_rate": 2.959110816468935e-07,
+      "loss": 0.3285,
+      "step": 847
+    },
+    {
+      "epoch": 2.70926517571885,
+      "grad_norm": 4.0700580829300685,
+      "learning_rate": 2.896434532852277e-07,
+      "loss": 0.3389,
+      "step": 848
+    },
+    {
+      "epoch": 2.7124600638977636,
+      "grad_norm": 3.6000635843692104,
+      "learning_rate": 2.834409337112842e-07,
+      "loss": 0.3371,
+      "step": 849
+    },
+    {
+      "epoch": 2.7156549520766773,
+      "grad_norm": 2.9152897449729664,
+      "learning_rate": 2.7730360865923954e-07,
+      "loss": 0.3096,
+      "step": 850
+    },
+    {
+      "epoch": 2.718849840255591,
+      "grad_norm": 3.0326316476814625,
+      "learning_rate": 2.712315629621176e-07,
+      "loss": 0.2754,
+      "step": 851
+    },
+    {
+      "epoch": 2.722044728434505,
+      "grad_norm": 2.7435229441530398,
+      "learning_rate": 2.6522488055062076e-07,
+      "loss": 0.3611,
+      "step": 852
+    },
+    {
+      "epoch": 2.7252396166134183,
+      "grad_norm": 2.324763442569114,
+      "learning_rate": 2.5928364445196975e-07,
+      "loss": 0.2833,
+      "step": 853
+    },
+    {
+      "epoch": 2.7284345047923324,
+      "grad_norm": 3.222681193417054,
+      "learning_rate": 2.534079367887549e-07,
+      "loss": 0.3757,
+      "step": 854
+    },
+    {
+      "epoch": 2.731629392971246,
+      "grad_norm": 2.6366157512843924,
+      "learning_rate": 2.475978387778e-07,
+      "loss": 0.272,
+      "step": 855
+    },
+    {
+      "epoch": 2.7348242811501597,
+      "grad_norm": 2.9546135711710244,
+      "learning_rate": 2.4185343072904376e-07,
+      "loss": 0.3972,
+      "step": 856
+    },
+    {
+      "epoch": 2.7380191693290734,
+      "grad_norm": 4.560718816586953,
+      "learning_rate": 2.3617479204442462e-07,
+      "loss": 0.3522,
+      "step": 857
+    },
+    {
+      "epoch": 2.741214057507987,
+      "grad_norm": 2.811195702355106,
+      "learning_rate": 2.305620012167853e-07,
+      "loss": 0.2764,
+      "step": 858
+    },
+    {
+      "epoch": 2.744408945686901,
+      "grad_norm": 2.7562929816171873,
+      "learning_rate": 2.2501513582879108e-07,
+      "loss": 0.2868,
+      "step": 859
+    },
+    {
+      "epoch": 2.747603833865815,
+      "grad_norm": 2.735901624101432,
+      "learning_rate": 2.1953427255185122e-07,
+      "loss": 0.2941,
+      "step": 860
+    },
+    {
+      "epoch": 2.7507987220447285,
+      "grad_norm": 3.462786776897098,
+      "learning_rate": 2.1411948714506414e-07,
+      "loss": 0.4092,
+      "step": 861
+    },
+    {
+      "epoch": 2.753993610223642,
+      "grad_norm": 2.3127819662003914,
+      "learning_rate": 2.0877085445416889e-07,
+      "loss": 0.2384,
+      "step": 862
+    },
+    {
+      "epoch": 2.757188498402556,
+      "grad_norm": 3.233248192138762,
+      "learning_rate": 2.034884484105093e-07,
+      "loss": 0.2754,
+      "step": 863
+    },
+    {
+      "epoch": 2.7603833865814695,
+      "grad_norm": 2.429855929992095,
+      "learning_rate": 1.98272342030012e-07,
+      "loss": 0.3137,
+      "step": 864
+    },
+    {
+      "epoch": 2.763578274760383,
+      "grad_norm": 3.3086564978960817,
+      "learning_rate": 1.9312260741218114e-07,
+      "loss": 0.3207,
+      "step": 865
+    },
+    {
+      "epoch": 2.7667731629392973,
+      "grad_norm": 3.200642690131128,
+      "learning_rate": 1.8803931573909584e-07,
+      "loss": 0.3023,
+      "step": 866
+    },
+    {
+      "epoch": 2.769968051118211,
+      "grad_norm": 2.658046632214419,
+      "learning_rate": 1.8302253727443041e-07,
+      "loss": 0.3366,
+      "step": 867
+    },
+    {
+      "epoch": 2.7731629392971247,
+      "grad_norm": 3.341162619560406,
+      "learning_rate": 1.7807234136248296e-07,
+      "loss": 0.4348,
+      "step": 868
+    },
+    {
+      "epoch": 2.7763578274760383,
+      "grad_norm": 2.6980152170410188,
+      "learning_rate": 1.731887964272144e-07,
+      "loss": 0.2885,
+      "step": 869
+    },
+    {
+      "epoch": 2.779552715654952,
+      "grad_norm": 2.8832765814188863,
+      "learning_rate": 1.6837196997130434e-07,
+      "loss": 0.4372,
+      "step": 870
+    },
+    {
+      "epoch": 2.7827476038338657,
+      "grad_norm": 3.0151079044769142,
+      "learning_rate": 1.6362192857521942e-07,
+      "loss": 0.2566,
+      "step": 871
+    },
+    {
+      "epoch": 2.7859424920127793,
+      "grad_norm": 2.6306899487136164,
+      "learning_rate": 1.5893873789628812e-07,
+      "loss": 0.3733,
+      "step": 872
+    },
+    {
+      "epoch": 2.7891373801916934,
+      "grad_norm": 2.8263310110158657,
+      "learning_rate": 1.5432246266780083e-07,
+      "loss": 0.3486,
+      "step": 873
+    },
+    {
+      "epoch": 2.792332268370607,
+      "grad_norm": 2.908318086441676,
+      "learning_rate": 1.4977316669810782e-07,
+      "loss": 0.2156,
+      "step": 874
+    },
+    {
+      "epoch": 2.7955271565495208,
+      "grad_norm": 3.088779492883142,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.4312,
+      "step": 875
+    },
+    {
+      "epoch": 2.7987220447284344,
+      "grad_norm": 3.3044228081886207,
+      "learning_rate": 1.4087576313854212e-07,
+      "loss": 0.4232,
+      "step": 876
+    },
+    {
+      "epoch": 2.801916932907348,
+      "grad_norm": 2.9321870387313322,
+      "learning_rate": 1.365277785328123e-07,
+      "loss": 0.3156,
+      "step": 877
+    },
+    {
+      "epoch": 2.8051118210862622,
+      "grad_norm": 4.142485710725183,
+      "learning_rate": 1.3224701915246053e-07,
+      "loss": 0.3429,
+      "step": 878
+    },
+    {
+      "epoch": 2.8083067092651754,
+      "grad_norm": 2.6058997907791572,
+      "learning_rate": 1.280335441681796e-07,
+      "loss": 0.2603,
+      "step": 879
+    },
+    {
+      "epoch": 2.8115015974440896,
+      "grad_norm": 2.657128150458897,
+      "learning_rate": 1.2388741182062348e-07,
+      "loss": 0.2914,
+      "step": 880
+    },
+    {
+      "epoch": 2.8146964856230032,
+      "grad_norm": 2.997770479140602,
+      "learning_rate": 1.198086794196035e-07,
+      "loss": 0.2696,
+      "step": 881
+    },
+    {
+      "epoch": 2.817891373801917,
+      "grad_norm": 2.72862453324739,
+      "learning_rate": 1.1579740334330014e-07,
+      "loss": 0.3682,
+      "step": 882
+    },
+    {
+      "epoch": 2.8210862619808306,
+      "grad_norm": 2.810387949044398,
+      "learning_rate": 1.1185363903747748e-07,
+      "loss": 0.3304,
+      "step": 883
+    },
+    {
+      "epoch": 2.8242811501597442,
+      "grad_norm": 3.4437166587675687,
+      "learning_rate": 1.0797744101472052e-07,
+      "loss": 0.3702,
+      "step": 884
+    },
+    {
+      "epoch": 2.8274760383386583,
+      "grad_norm": 2.507561319593953,
+      "learning_rate": 1.0416886285368188e-07,
+      "loss": 0.2157,
+      "step": 885
+    },
+    {
+      "epoch": 2.830670926517572,
+      "grad_norm": 2.5438117608192825,
+      "learning_rate": 1.0042795719833964e-07,
+      "loss": 0.2359,
+      "step": 886
+    },
+    {
+      "epoch": 2.8338658146964857,
+      "grad_norm": 3.3080500710422323,
+      "learning_rate": 9.675477575726954e-08,
+      "loss": 0.4275,
+      "step": 887
+    },
+    {
+      "epoch": 2.8370607028753994,
+      "grad_norm": 2.9036884561997094,
+      "learning_rate": 9.314936930293283e-08,
+      "loss": 0.2833,
+      "step": 888
+    },
+    {
+      "epoch": 2.840255591054313,
+      "grad_norm": 2.842119513399336,
+      "learning_rate": 8.961178767097178e-08,
+      "loss": 0.3865,
+      "step": 889
+    },
+    {
+      "epoch": 2.8434504792332267,
+      "grad_norm": 2.9159590616354207,
+      "learning_rate": 8.614207975952083e-08,
+      "loss": 0.2992,
+      "step": 890
+    },
+    {
+      "epoch": 2.8466453674121404,
+      "grad_norm": 2.867001302771436,
+      "learning_rate": 8.274029352853264e-08,
+      "loss": 0.2968,
+      "step": 891
+    },
+    {
+      "epoch": 2.8498402555910545,
+      "grad_norm": 2.6724058618221997,
+      "learning_rate": 7.940647599911477e-08,
+      "loss": 0.2731,
+      "step": 892
+    },
+    {
+      "epoch": 2.853035143769968,
+      "grad_norm": 2.8207778080885766,
+      "learning_rate": 7.614067325287632e-08,
+      "loss": 0.3856,
+      "step": 893
+    },
+    {
+      "epoch": 2.856230031948882,
+      "grad_norm": 4.062544990755319,
+      "learning_rate": 7.294293043129785e-08,
+      "loss": 0.3013,
+      "step": 894
+    },
+    {
+      "epoch": 2.8594249201277955,
+      "grad_norm": 3.085694435173444,
+      "learning_rate": 6.981329173509909e-08,
+      "loss": 0.3666,
+      "step": 895
+    },
+    {
+      "epoch": 2.862619808306709,
+      "grad_norm": 2.8582415008994144,
+      "learning_rate": 6.675180042363505e-08,
+      "loss": 0.3128,
+      "step": 896
+    },
+    {
+      "epoch": 2.8658146964856233,
+      "grad_norm": 3.179282362572499,
+      "learning_rate": 6.375849881429418e-08,
+      "loss": 0.3061,
+      "step": 897
+    },
+    {
+      "epoch": 2.8690095846645365,
+      "grad_norm": 2.5115716965563424,
+      "learning_rate": 6.083342828191453e-08,
+      "loss": 0.2858,
+      "step": 898
+    },
+    {
+      "epoch": 2.8722044728434506,
+      "grad_norm": 2.777401669431884,
+      "learning_rate": 5.797662925821068e-08,
+      "loss": 0.3512,
+      "step": 899
+    },
+    {
+      "epoch": 2.8753993610223643,
+      "grad_norm": 2.9117302321232925,
+      "learning_rate": 5.518814123121885e-08,
+      "loss": 0.2598,
+      "step": 900
+    },
+    {
+      "epoch": 2.878594249201278,
+      "grad_norm": 2.678128871907406,
+      "learning_rate": 5.246800274474439e-08,
+      "loss": 0.3006,
+      "step": 901
+    },
+    {
+      "epoch": 2.8817891373801916,
+      "grad_norm": 2.663890443758662,
+      "learning_rate": 4.981625139783619e-08,
+      "loss": 0.2548,
+      "step": 902
+    },
+    {
+      "epoch": 2.8849840255591053,
+      "grad_norm": 2.7167809056917966,
+      "learning_rate": 4.723292384426203e-08,
+      "loss": 0.2704,
+      "step": 903
+    },
+    {
+      "epoch": 2.8881789137380194,
+      "grad_norm": 2.851329777902932,
+      "learning_rate": 4.471805579200239e-08,
+      "loss": 0.2287,
+      "step": 904
+    },
+    {
+      "epoch": 2.891373801916933,
+      "grad_norm": 2.621609109582403,
+      "learning_rate": 4.227168200276077e-08,
+      "loss": 0.2758,
+      "step": 905
+    },
+    {
+      "epoch": 2.8945686900958467,
+      "grad_norm": 2.728319696189122,
+      "learning_rate": 3.989383629147747e-08,
+      "loss": 0.3029,
+      "step": 906
+    },
+    {
+      "epoch": 2.8977635782747604,
+      "grad_norm": 2.384165536174641,
+      "learning_rate": 3.758455152586715e-08,
+      "loss": 0.278,
+      "step": 907
+    },
+    {
+      "epoch": 2.900958466453674,
+      "grad_norm": 3.045420615213383,
+      "learning_rate": 3.534385962596143e-08,
+      "loss": 0.269,
+      "step": 908
+    },
+    {
+      "epoch": 2.9041533546325877,
+      "grad_norm": 3.5274917559035375,
+      "learning_rate": 3.3171791563669785e-08,
+      "loss": 0.3614,
+      "step": 909
+    },
+    {
+      "epoch": 2.9073482428115014,
+      "grad_norm": 3.037022122609774,
+      "learning_rate": 3.10683773623488e-08,
+      "loss": 0.3431,
+      "step": 910
+    },
+    {
+      "epoch": 2.9105431309904155,
+      "grad_norm": 2.3606663319025123,
+      "learning_rate": 2.9033646096390255e-08,
+      "loss": 0.1731,
+      "step": 911
+    },
+    {
+      "epoch": 2.913738019169329,
+      "grad_norm": 2.7643789730096233,
+      "learning_rate": 2.706762589081646e-08,
+      "loss": 0.2904,
+      "step": 912
+    },
+    {
+      "epoch": 2.916932907348243,
+      "grad_norm": 2.8040995445734533,
+      "learning_rate": 2.517034392089446e-08,
+      "loss": 0.3038,
+      "step": 913
+    },
+    {
+      "epoch": 2.9201277955271565,
+      "grad_norm": 2.6135234405506247,
+      "learning_rate": 2.3341826411756863e-08,
+      "loss": 0.2799,
+      "step": 914
+    },
+    {
+      "epoch": 2.92332268370607,
+      "grad_norm": 2.7401452742324266,
+      "learning_rate": 2.158209863804217e-08,
+      "loss": 0.2882,
+      "step": 915
+    },
+    {
+      "epoch": 2.9265175718849843,
+      "grad_norm": 2.9820183469996593,
+      "learning_rate": 1.9891184923544472e-08,
+      "loss": 0.3816,
+      "step": 916
+    },
+    {
+      "epoch": 2.9297124600638975,
+      "grad_norm": 3.0533718377928913,
+      "learning_rate": 1.826910864087761e-08,
+      "loss": 0.4242,
+      "step": 917
+    },
+    {
+      "epoch": 2.9329073482428116,
+      "grad_norm": 2.9581401477503917,
+      "learning_rate": 1.6715892211150442e-08,
+      "loss": 0.3372,
+      "step": 918
+    },
+    {
+      "epoch": 2.9361022364217253,
+      "grad_norm": 2.8361015784905312,
+      "learning_rate": 1.5231557103658755e-08,
+      "loss": 0.4192,
+      "step": 919
+    },
+    {
+      "epoch": 2.939297124600639,
+      "grad_norm": 3.0586426730909655,
+      "learning_rate": 1.3816123835588835e-08,
+      "loss": 0.3581,
+      "step": 920
+    },
+    {
+      "epoch": 2.9424920127795526,
+      "grad_norm": 2.5427136786061344,
+      "learning_rate": 1.2469611971731576e-08,
+      "loss": 0.2739,
+      "step": 921
+    },
+    {
+      "epoch": 2.9456869009584663,
+      "grad_norm": 2.1712667914026027,
+      "learning_rate": 1.1192040124214931e-08,
+      "loss": 0.2326,
+      "step": 922
+    },
+    {
+      "epoch": 2.9488817891373804,
+      "grad_norm": 3.188485491426543,
+      "learning_rate": 9.983425952243552e-09,
+      "loss": 0.3922,
+      "step": 923
+    },
+    {
+      "epoch": 2.952076677316294,
+      "grad_norm": 2.845843916729636,
+      "learning_rate": 8.84378616185788e-09,
+      "loss": 0.3286,
+      "step": 924
+    },
+    {
+      "epoch": 2.9552715654952078,
+      "grad_norm": 3.5765032411749336,
+      "learning_rate": 7.773136505700995e-09,
+      "loss": 0.331,
+      "step": 925
+    },
+    {
+      "epoch": 2.9584664536741214,
+      "grad_norm": 2.8788159442183456,
+      "learning_rate": 6.7714917828004545e-09,
+      "loss": 0.3166,
+      "step": 926
+    },
+    {
+      "epoch": 2.961661341853035,
+      "grad_norm": 2.741962504064861,
+      "learning_rate": 5.838865838366792e-09,
+      "loss": 0.4507,
+      "step": 927
+    },
+    {
+      "epoch": 2.9648562300319488,
+      "grad_norm": 3.2366842938340996,
+      "learning_rate": 4.975271563599227e-09,
+      "loss": 0.302,
+      "step": 928
+    },
+    {
+      "epoch": 2.9680511182108624,
+      "grad_norm": 2.3623682591288695,
+      "learning_rate": 4.180720895508028e-09,
+      "loss": 0.3055,
+      "step": 929
+    },
+    {
+      "epoch": 2.9712460063897765,
+      "grad_norm": 2.806148591640003,
+      "learning_rate": 3.4552248167507576e-09,
+      "loss": 0.3891,
+      "step": 930
+    },
+    {
+      "epoch": 2.97444089456869,
+      "grad_norm": 2.3827046811960937,
+      "learning_rate": 2.798793355478502e-09,
+      "loss": 0.2795,
+      "step": 931
+    },
+    {
+      "epoch": 2.977635782747604,
+      "grad_norm": 3.5169477302486114,
+      "learning_rate": 2.2114355851993175e-09,
+      "loss": 0.3361,
+      "step": 932
+    },
+    {
+      "epoch": 2.9808306709265175,
+      "grad_norm": 2.595128104000001,
+      "learning_rate": 1.6931596246516636e-09,
+      "loss": 0.2227,
+      "step": 933
+    },
+    {
+      "epoch": 2.984025559105431,
+      "grad_norm": 3.143480292395146,
+      "learning_rate": 1.24397263769227e-09,
+      "loss": 0.3025,
+      "step": 934
+    },
+    {
+      "epoch": 2.987220447284345,
+      "grad_norm": 2.6724149710186795,
+      "learning_rate": 8.638808331973281e-10,
+      "loss": 0.3166,
+      "step": 935
+    },
+    {
+      "epoch": 2.9904153354632586,
+      "grad_norm": 2.9142255522023985,
+      "learning_rate": 5.528894649758921e-10,
+      "loss": 0.3725,
+      "step": 936
+    },
+    {
+      "epoch": 2.9936102236421727,
+      "grad_norm": 2.7175802933069226,
+      "learning_rate": 3.1100283169938074e-10,
+      "loss": 0.343,
+      "step": 937
+    },
+    {
+      "epoch": 2.9968051118210863,
+      "grad_norm": 3.325209591123207,
+      "learning_rate": 1.3822427683884975e-10,
+      "loss": 0.3749,
+      "step": 938
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.3936701387445036,
+      "learning_rate": 3.4556188622802964e-11,
+      "loss": 0.1691,
+      "step": 939
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 939,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5691721383936.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}