diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24921 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8948781386534717,
+  "eval_steps": 500,
+  "global_step": 68000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002631994525451387,
+      "grad_norm": 3.5187880992889404,
+      "learning_rate": 8.771929824561404e-08,
+      "loss": 0.5368,
+      "step": 20
+    },
+    {
+      "epoch": 0.0005263989050902774,
+      "grad_norm": 4.334347248077393,
+      "learning_rate": 1.7543859649122808e-07,
+      "loss": 0.4929,
+      "step": 40
+    },
+    {
+      "epoch": 0.0007895983576354162,
+      "grad_norm": 5.199437618255615,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 0.4515,
+      "step": 60
+    },
+    {
+      "epoch": 0.0010527978101805549,
+      "grad_norm": 4.703293323516846,
+      "learning_rate": 3.5087719298245616e-07,
+      "loss": 0.4764,
+      "step": 80
+    },
+    {
+      "epoch": 0.0013159972627256936,
+      "grad_norm": 2.2297756671905518,
+      "learning_rate": 4.385964912280702e-07,
+      "loss": 0.4647,
+      "step": 100
+    },
+    {
+      "epoch": 0.0015791967152708323,
+      "grad_norm": 3.402606964111328,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 0.467,
+      "step": 120
+    },
+    {
+      "epoch": 0.001842396167815971,
+      "grad_norm": 4.127569198608398,
+      "learning_rate": 6.140350877192982e-07,
+      "loss": 0.3906,
+      "step": 140
+    },
+    {
+      "epoch": 0.0021055956203611097,
+      "grad_norm": 1.7735623121261597,
+      "learning_rate": 7.017543859649123e-07,
+      "loss": 0.4365,
+      "step": 160
+    },
+    {
+      "epoch": 0.0023687950729062482,
+      "grad_norm": 1.766693353652954,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 0.3554,
+      "step": 180
+    },
+    {
+      "epoch": 0.002631994525451387,
+      "grad_norm": 1.3966196775436401,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 0.3673,
+      "step": 200
+    },
+    {
+      "epoch": 0.0028951939779965257,
+      "grad_norm": 1.7526930570602417,
+      "learning_rate": 9.649122807017545e-07,
+      "loss": 0.3487,
+      "step": 220
+    },
+    {
+      "epoch": 0.0031583934305416646,
+      "grad_norm": 2.094742774963379,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 0.3486,
+      "step": 240
+    },
+    {
+      "epoch": 0.003421592883086803,
+      "grad_norm": 1.5583423376083374,
+      "learning_rate": 1.1403508771929824e-06,
+      "loss": 0.3224,
+      "step": 260
+    },
+    {
+      "epoch": 0.003684792335631942,
+      "grad_norm": 1.7672529220581055,
+      "learning_rate": 1.2280701754385965e-06,
+      "loss": 0.3288,
+      "step": 280
+    },
+    {
+      "epoch": 0.0039479917881770805,
+      "grad_norm": 1.304077386856079,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.3057,
+      "step": 300
+    },
+    {
+      "epoch": 0.0042111912407222195,
+      "grad_norm": 2.1879703998565674,
+      "learning_rate": 1.4035087719298246e-06,
+      "loss": 0.3081,
+      "step": 320
+    },
+    {
+      "epoch": 0.004474390693267358,
+      "grad_norm": 4.220763206481934,
+      "learning_rate": 1.4912280701754387e-06,
+      "loss": 0.2853,
+      "step": 340
+    },
+    {
+      "epoch": 0.0047375901458124965,
+      "grad_norm": 1.9660420417785645,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 0.2935,
+      "step": 360
+    },
+    {
+      "epoch": 0.005000789598357635,
+      "grad_norm": 2.218201160430908,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.3147,
+      "step": 380
+    },
+    {
+      "epoch": 0.005263989050902774,
+      "grad_norm": 2.7722785472869873,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 0.3041,
+      "step": 400
+    },
+    {
+      "epoch": 0.005527188503447912,
+      "grad_norm": 1.3512650728225708,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 0.29,
+      "step": 420
+    },
+    {
+      "epoch": 0.005790387955993051,
+      "grad_norm": 1.9269977807998657,
+      "learning_rate": 1.929824561403509e-06,
+      "loss": 0.2994,
+      "step": 440
+    },
+    {
+      "epoch": 0.00605358740853819,
+      "grad_norm": 1.4007967710494995,
+      "learning_rate": 2.017543859649123e-06,
+      "loss": 0.2678,
+      "step": 460
+    },
+    {
+      "epoch": 0.006316786861083329,
+      "grad_norm": 1.5149136781692505,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 0.2634,
+      "step": 480
+    },
+    {
+      "epoch": 0.006579986313628467,
+      "grad_norm": 1.4462240934371948,
+      "learning_rate": 2.192982456140351e-06,
+      "loss": 0.2745,
+      "step": 500
+    },
+    {
+      "epoch": 0.006579986313628467,
+      "eval_loss": 0.3154577612876892,
+      "eval_runtime": 4.8137,
+      "eval_samples_per_second": 3.324,
+      "eval_steps_per_second": 1.662,
+      "step": 500
+    },
+    {
+      "epoch": 0.006843185766173606,
+      "grad_norm": 1.5516880750656128,
+      "learning_rate": 2.280701754385965e-06,
+      "loss": 0.2678,
+      "step": 520
+    },
+    {
+      "epoch": 0.007106385218718745,
+      "grad_norm": 1.6810089349746704,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 0.2821,
+      "step": 540
+    },
+    {
+      "epoch": 0.007369584671263884,
+      "grad_norm": 2.0241265296936035,
+      "learning_rate": 2.456140350877193e-06,
+      "loss": 0.2818,
+      "step": 560
+    },
+    {
+      "epoch": 0.007632784123809022,
+      "grad_norm": 1.6910285949707031,
+      "learning_rate": 2.5438596491228075e-06,
+      "loss": 0.2438,
+      "step": 580
+    },
+    {
+      "epoch": 0.007895983576354161,
+      "grad_norm": 1.9050703048706055,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.2814,
+      "step": 600
+    },
+    {
+      "epoch": 0.0081591830288993,
+      "grad_norm": 1.3354547023773193,
+      "learning_rate": 2.7192982456140356e-06,
+      "loss": 0.255,
+      "step": 620
+    },
+    {
+      "epoch": 0.008422382481444439,
+      "grad_norm": 1.5497230291366577,
+      "learning_rate": 2.8070175438596493e-06,
+      "loss": 0.2634,
+      "step": 640
+    },
+    {
+      "epoch": 0.008685581933989578,
+      "grad_norm": 2.065305233001709,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 0.267,
+      "step": 660
+    },
+    {
+      "epoch": 0.008948781386534717,
+      "grad_norm": 1.4903022050857544,
+      "learning_rate": 2.9824561403508774e-06,
+      "loss": 0.2576,
+      "step": 680
+    },
+    {
+      "epoch": 0.009211980839079854,
+      "grad_norm": 2.1075193881988525,
+      "learning_rate": 3.0701754385964915e-06,
+      "loss": 0.2617,
+      "step": 700
+    },
+    {
+      "epoch": 0.009475180291624993,
+      "grad_norm": 2.1362106800079346,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.2689,
+      "step": 720
+    },
+    {
+      "epoch": 0.009738379744170132,
+      "grad_norm": 1.677917718887329,
+      "learning_rate": 3.2456140350877197e-06,
+      "loss": 0.2508,
+      "step": 740
+    },
+    {
+      "epoch": 0.01000157919671527,
+      "grad_norm": 1.7670167684555054,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2516,
+      "step": 760
+    },
+    {
+      "epoch": 0.01026477864926041,
+      "grad_norm": 1.8894158601760864,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 0.2499,
+      "step": 780
+    },
+    {
+      "epoch": 0.010527978101805549,
+      "grad_norm": 1.4801076650619507,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 0.238,
+      "step": 800
+    },
+    {
+      "epoch": 0.010791177554350688,
+      "grad_norm": 1.6919801235198975,
+      "learning_rate": 3.596491228070176e-06,
+      "loss": 0.2547,
+      "step": 820
+    },
+    {
+      "epoch": 0.011054377006895825,
+      "grad_norm": 2.0446670055389404,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 0.2405,
+      "step": 840
+    },
+    {
+      "epoch": 0.011317576459440964,
+      "grad_norm": 1.864290714263916,
+      "learning_rate": 3.7719298245614037e-06,
+      "loss": 0.2322,
+      "step": 860
+    },
+    {
+      "epoch": 0.011580775911986103,
+      "grad_norm": 1.3105003833770752,
+      "learning_rate": 3.859649122807018e-06,
+      "loss": 0.2558,
+      "step": 880
+    },
+    {
+      "epoch": 0.011843975364531242,
+      "grad_norm": 1.4571141004562378,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 0.2236,
+      "step": 900
+    },
+    {
+      "epoch": 0.01210717481707638,
+      "grad_norm": 1.8333299160003662,
+      "learning_rate": 4.035087719298246e-06,
+      "loss": 0.2549,
+      "step": 920
+    },
+    {
+      "epoch": 0.01237037426962152,
+      "grad_norm": 1.418653964996338,
+      "learning_rate": 4.12280701754386e-06,
+      "loss": 0.2489,
+      "step": 940
+    },
+    {
+      "epoch": 0.012633573722166658,
+      "grad_norm": 1.3325315713882446,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.2558,
+      "step": 960
+    },
+    {
+      "epoch": 0.012896773174711797,
+      "grad_norm": 2.4286279678344727,
+      "learning_rate": 4.298245614035088e-06,
+      "loss": 0.2476,
+      "step": 980
+    },
+    {
+      "epoch": 0.013159972627256935,
+      "grad_norm": 1.8234083652496338,
+      "learning_rate": 4.385964912280702e-06,
+      "loss": 0.2362,
+      "step": 1000
+    },
+    {
+      "epoch": 0.013159972627256935,
+      "eval_loss": 0.27868399024009705,
+      "eval_runtime": 4.3016,
+      "eval_samples_per_second": 3.72,
+      "eval_steps_per_second": 1.86,
+      "step": 1000
+    },
+    {
+      "epoch": 0.013423172079802074,
+      "grad_norm": 1.054779291152954,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 0.2572,
+      "step": 1020
+    },
+    {
+      "epoch": 0.013686371532347212,
+      "grad_norm": 1.702553153038025,
+      "learning_rate": 4.56140350877193e-06,
+      "loss": 0.2529,
+      "step": 1040
+    },
+    {
+      "epoch": 0.013949570984892351,
+      "grad_norm": 1.6957552433013916,
+      "learning_rate": 4.649122807017544e-06,
+      "loss": 0.2387,
+      "step": 1060
+    },
+    {
+      "epoch": 0.01421277043743749,
+      "grad_norm": 1.0628148317337036,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.2469,
+      "step": 1080
+    },
+    {
+      "epoch": 0.01447596988998263,
+      "grad_norm": 1.645086407661438,
+      "learning_rate": 4.824561403508772e-06,
+      "loss": 0.2548,
+      "step": 1100
+    },
+    {
+      "epoch": 0.014739169342527768,
+      "grad_norm": 1.6618074178695679,
+      "learning_rate": 4.912280701754386e-06,
+      "loss": 0.2462,
+      "step": 1120
+    },
+    {
+      "epoch": 0.015002368795072905,
+      "grad_norm": 1.3233495950698853,
+      "learning_rate": 5e-06,
+      "loss": 0.246,
+      "step": 1140
+    },
+    {
+      "epoch": 0.015265568247618044,
+      "grad_norm": 1.3222942352294922,
+      "learning_rate": 5.087719298245615e-06,
+      "loss": 0.2544,
+      "step": 1160
+    },
+    {
+      "epoch": 0.015528767700163183,
+      "grad_norm": 1.4435482025146484,
+      "learning_rate": 5.175438596491229e-06,
+      "loss": 0.2403,
+      "step": 1180
+    },
+    {
+      "epoch": 0.015791967152708322,
+      "grad_norm": 1.5912235975265503,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.235,
+      "step": 1200
+    },
+    {
+      "epoch": 0.01605516660525346,
+      "grad_norm": 1.3008149862289429,
+      "learning_rate": 5.350877192982457e-06,
+      "loss": 0.2392,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0163183660577986,
+      "grad_norm": 1.5564367771148682,
+      "learning_rate": 5.438596491228071e-06,
+      "loss": 0.2549,
+      "step": 1240
+    },
+    {
+      "epoch": 0.01658156551034374,
+      "grad_norm": 1.4823662042617798,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 0.2279,
+      "step": 1260
+    },
+    {
+      "epoch": 0.016844764962888878,
+      "grad_norm": 1.501945972442627,
+      "learning_rate": 5.6140350877192985e-06,
+      "loss": 0.2244,
+      "step": 1280
+    },
+    {
+      "epoch": 0.017107964415434017,
+      "grad_norm": 1.7196705341339111,
+      "learning_rate": 5.701754385964913e-06,
+      "loss": 0.2505,
+      "step": 1300
+    },
+    {
+      "epoch": 0.017371163867979156,
+      "grad_norm": 1.147748589515686,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 0.2339,
+      "step": 1320
+    },
+    {
+      "epoch": 0.017634363320524295,
+      "grad_norm": 1.4615315198898315,
+      "learning_rate": 5.877192982456141e-06,
+      "loss": 0.2361,
+      "step": 1340
+    },
+    {
+      "epoch": 0.017897562773069434,
+      "grad_norm": 1.3070178031921387,
+      "learning_rate": 5.964912280701755e-06,
+      "loss": 0.2332,
+      "step": 1360
+    },
+    {
+      "epoch": 0.01816076222561457,
+      "grad_norm": 2.078519582748413,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 0.2254,
+      "step": 1380
+    },
+    {
+      "epoch": 0.018423961678159708,
+      "grad_norm": 1.7956151962280273,
+      "learning_rate": 6.140350877192983e-06,
+      "loss": 0.2332,
+      "step": 1400
+    },
+    {
+      "epoch": 0.018687161130704847,
+      "grad_norm": 2.05875301361084,
+      "learning_rate": 6.2280701754385975e-06,
+      "loss": 0.2543,
+      "step": 1420
+    },
+    {
+      "epoch": 0.018950360583249986,
+      "grad_norm": 1.874975323677063,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 0.2251,
+      "step": 1440
+    },
+    {
+      "epoch": 0.019213560035795125,
+      "grad_norm": 1.228518009185791,
+      "learning_rate": 6.403508771929825e-06,
+      "loss": 0.2445,
+      "step": 1460
+    },
+    {
+      "epoch": 0.019476759488340264,
+      "grad_norm": 1.9091897010803223,
+      "learning_rate": 6.491228070175439e-06,
+      "loss": 0.2172,
+      "step": 1480
+    },
+    {
+      "epoch": 0.019739958940885403,
+      "grad_norm": 1.716107964515686,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 0.2184,
+      "step": 1500
+    },
+    {
+      "epoch": 0.019739958940885403,
+      "eval_loss": 0.26523810625076294,
+      "eval_runtime": 4.1583,
+      "eval_samples_per_second": 3.848,
+      "eval_steps_per_second": 1.924,
+      "step": 1500
+    },
+    {
+      "epoch": 0.02000315839343054,
+      "grad_norm": 1.6500223875045776,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.2323,
+      "step": 1520
+    },
+    {
+      "epoch": 0.02026635784597568,
+      "grad_norm": 1.1524165868759155,
+      "learning_rate": 6.754385964912281e-06,
+      "loss": 0.2245,
+      "step": 1540
+    },
+    {
+      "epoch": 0.02052955729852082,
+      "grad_norm": 1.4364941120147705,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.2462,
+      "step": 1560
+    },
+    {
+      "epoch": 0.02079275675106596,
+      "grad_norm": 1.9232507944107056,
+      "learning_rate": 6.92982456140351e-06,
+      "loss": 0.2278,
+      "step": 1580
+    },
+    {
+      "epoch": 0.021055956203611097,
+      "grad_norm": 1.786876916885376,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 0.2463,
+      "step": 1600
+    },
+    {
+      "epoch": 0.021319155656156236,
+      "grad_norm": 1.5189566612243652,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 0.2461,
+      "step": 1620
+    },
+    {
+      "epoch": 0.021582355108701375,
+      "grad_norm": 1.9569761753082275,
+      "learning_rate": 7.192982456140352e-06,
+      "loss": 0.2182,
+      "step": 1640
+    },
+    {
+      "epoch": 0.021845554561246514,
+      "grad_norm": 1.57135808467865,
+      "learning_rate": 7.280701754385966e-06,
+      "loss": 0.2456,
+      "step": 1660
+    },
+    {
+      "epoch": 0.02210875401379165,
+      "grad_norm": 1.2492105960845947,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.239,
+      "step": 1680
+    },
+    {
+      "epoch": 0.02237195346633679,
+      "grad_norm": 1.8191033601760864,
+      "learning_rate": 7.456140350877194e-06,
+      "loss": 0.2316,
+      "step": 1700
+    },
+    {
+      "epoch": 0.022635152918881928,
+      "grad_norm": 1.7759687900543213,
+      "learning_rate": 7.5438596491228074e-06,
+      "loss": 0.2419,
+      "step": 1720
+    },
+    {
+      "epoch": 0.022898352371427066,
+      "grad_norm": 1.2422982454299927,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 0.2241,
+      "step": 1740
+    },
+    {
+      "epoch": 0.023161551823972205,
+      "grad_norm": 1.7862449884414673,
+      "learning_rate": 7.719298245614036e-06,
+      "loss": 0.2338,
+      "step": 1760
+    },
+    {
+      "epoch": 0.023424751276517344,
+      "grad_norm": 1.288135290145874,
+      "learning_rate": 7.80701754385965e-06,
+      "loss": 0.2412,
+      "step": 1780
+    },
+    {
+      "epoch": 0.023687950729062483,
+      "grad_norm": 1.4992725849151611,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.2335,
+      "step": 1800
+    },
+    {
+      "epoch": 0.023951150181607622,
+      "grad_norm": 1.4365191459655762,
+      "learning_rate": 7.982456140350877e-06,
+      "loss": 0.2282,
+      "step": 1820
+    },
+    {
+      "epoch": 0.02421434963415276,
+      "grad_norm": 1.4234129190444946,
+      "learning_rate": 8.070175438596492e-06,
+      "loss": 0.2456,
+      "step": 1840
+    },
+    {
+      "epoch": 0.0244775490866979,
+      "grad_norm": 1.1815886497497559,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 0.2386,
+      "step": 1860
+    },
+    {
+      "epoch": 0.02474074853924304,
+      "grad_norm": 1.9235172271728516,
+      "learning_rate": 8.24561403508772e-06,
+      "loss": 0.2235,
+      "step": 1880
+    },
+    {
+      "epoch": 0.025003947991788178,
+      "grad_norm": 1.756247639656067,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2383,
+      "step": 1900
+    },
+    {
+      "epoch": 0.025267147444333317,
+      "grad_norm": 2.9298629760742188,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.2288,
+      "step": 1920
+    },
+    {
+      "epoch": 0.025530346896878456,
+      "grad_norm": 1.1361719369888306,
+      "learning_rate": 8.508771929824563e-06,
+      "loss": 0.2363,
+      "step": 1940
+    },
+    {
+      "epoch": 0.025793546349423595,
+      "grad_norm": 1.8624403476715088,
+      "learning_rate": 8.596491228070176e-06,
+      "loss": 0.2353,
+      "step": 1960
+    },
+    {
+      "epoch": 0.02605674580196873,
+      "grad_norm": 1.9579325914382935,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 0.2388,
+      "step": 1980
+    },
+    {
+      "epoch": 0.02631994525451387,
+      "grad_norm": 1.4603593349456787,
+      "learning_rate": 8.771929824561405e-06,
+      "loss": 0.2256,
+      "step": 2000
+    },
+    {
+      "epoch": 0.02631994525451387,
+      "eval_loss": 0.2594212591648102,
+      "eval_runtime": 4.2968,
+      "eval_samples_per_second": 3.724,
+      "eval_steps_per_second": 1.862,
+      "step": 2000
+    },
+    {
+      "epoch": 0.026583144707059008,
+      "grad_norm": 1.7783149480819702,
+      "learning_rate": 8.859649122807017e-06,
+      "loss": 0.2496,
+      "step": 2020
+    },
+    {
+      "epoch": 0.026846344159604147,
+      "grad_norm": 1.6160141229629517,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.2506,
+      "step": 2040
+    },
+    {
+      "epoch": 0.027109543612149286,
+      "grad_norm": 1.323948621749878,
+      "learning_rate": 9.035087719298246e-06,
+      "loss": 0.2556,
+      "step": 2060
+    },
+    {
+      "epoch": 0.027372743064694425,
+      "grad_norm": 1.0840163230895996,
+      "learning_rate": 9.12280701754386e-06,
+      "loss": 0.2112,
+      "step": 2080
+    },
+    {
+      "epoch": 0.027635942517239564,
+      "grad_norm": 1.5541092157363892,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 0.224,
+      "step": 2100
+    },
+    {
+      "epoch": 0.027899141969784703,
+      "grad_norm": 1.6916871070861816,
+      "learning_rate": 9.298245614035088e-06,
+      "loss": 0.2409,
+      "step": 2120
+    },
+    {
+      "epoch": 0.02816234142232984,
+      "grad_norm": 1.422555923461914,
+      "learning_rate": 9.385964912280703e-06,
+      "loss": 0.2538,
+      "step": 2140
+    },
+    {
+      "epoch": 0.02842554087487498,
+      "grad_norm": 1.8744568824768066,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.2465,
+      "step": 2160
+    },
+    {
+      "epoch": 0.02868874032742012,
+      "grad_norm": 1.5689542293548584,
+      "learning_rate": 9.56140350877193e-06,
+      "loss": 0.2428,
+      "step": 2180
+    },
+    {
+      "epoch": 0.02895193977996526,
+      "grad_norm": 1.5494287014007568,
+      "learning_rate": 9.649122807017545e-06,
+      "loss": 0.2417,
+      "step": 2200
+    },
+    {
+      "epoch": 0.029215139232510397,
+      "grad_norm": 1.4703900814056396,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 0.2213,
+      "step": 2220
+    },
+    {
+      "epoch": 0.029478338685055536,
+      "grad_norm": 2.0383410453796387,
+      "learning_rate": 9.824561403508772e-06,
+      "loss": 0.2306,
+      "step": 2240
+    },
+    {
+      "epoch": 0.029741538137600675,
+      "grad_norm": 1.6995047330856323,
+      "learning_rate": 9.912280701754386e-06,
+      "loss": 0.2521,
+      "step": 2260
+    },
+    {
+      "epoch": 0.03000473759014581,
+      "grad_norm": 1.8226897716522217,
+      "learning_rate": 1e-05,
+      "loss": 0.2391,
+      "step": 2280
+    },
+    {
+      "epoch": 0.03026793704269095,
+      "grad_norm": 1.2334691286087036,
+      "learning_rate": 9.999998183353341e-06,
+      "loss": 0.2455,
+      "step": 2300
+    },
+    {
+      "epoch": 0.03053113649523609,
+      "grad_norm": 1.5180833339691162,
+      "learning_rate": 9.99999273341468e-06,
+      "loss": 0.2291,
+      "step": 2320
+    },
+    {
+      "epoch": 0.030794335947781228,
+      "grad_norm": 1.9181222915649414,
+      "learning_rate": 9.99998365018798e-06,
+      "loss": 0.2179,
+      "step": 2340
+    },
+    {
+      "epoch": 0.031057535400326367,
+      "grad_norm": 1.7804433107376099,
+      "learning_rate": 9.99997093367984e-06,
+      "loss": 0.238,
+      "step": 2360
+    },
+    {
+      "epoch": 0.031320734852871505,
+      "grad_norm": 1.5217117071151733,
+      "learning_rate": 9.9999545838995e-06,
+      "loss": 0.2417,
+      "step": 2380
+    },
+    {
+      "epoch": 0.031583934305416644,
+      "grad_norm": 1.7381410598754883,
+      "learning_rate": 9.999934600858846e-06,
+      "loss": 0.2355,
+      "step": 2400
+    },
+    {
+      "epoch": 0.03184713375796178,
+      "grad_norm": 1.5798227787017822,
+      "learning_rate": 9.99991098457239e-06,
+      "loss": 0.2193,
+      "step": 2420
+    },
+    {
+      "epoch": 0.03211033321050692,
+      "grad_norm": 1.9656158685684204,
+      "learning_rate": 9.999883735057298e-06,
+      "loss": 0.2224,
+      "step": 2440
+    },
+    {
+      "epoch": 0.03237353266305206,
+      "grad_norm": 1.5060995817184448,
+      "learning_rate": 9.999852852333372e-06,
+      "loss": 0.2193,
+      "step": 2460
+    },
+    {
+      "epoch": 0.0326367321155972,
+      "grad_norm": 1.5659561157226562,
+      "learning_rate": 9.999818336423051e-06,
+      "loss": 0.2288,
+      "step": 2480
+    },
+    {
+      "epoch": 0.03289993156814234,
+      "grad_norm": 1.4684646129608154,
+      "learning_rate": 9.999780187351418e-06,
+      "loss": 0.24,
+      "step": 2500
+    },
+    {
+      "epoch": 0.03289993156814234,
+      "eval_loss": 0.2620845437049866,
+      "eval_runtime": 3.5318,
+      "eval_samples_per_second": 4.53,
+      "eval_steps_per_second": 2.265,
+      "step": 2500
+    },
+    {
+      "epoch": 0.03316313102068748,
+      "grad_norm": 1.3452467918395996,
+      "learning_rate": 9.999738405146193e-06,
+      "loss": 0.2303,
+      "step": 2520
+    },
+    {
+      "epoch": 0.03342633047323262,
+      "grad_norm": 1.69887113571167,
+      "learning_rate": 9.999692989837738e-06,
+      "loss": 0.2272,
+      "step": 2540
+    },
+    {
+      "epoch": 0.033689529925777756,
+      "grad_norm": 1.8199821710586548,
+      "learning_rate": 9.999643941459055e-06,
+      "loss": 0.2377,
+      "step": 2560
+    },
+    {
+      "epoch": 0.033952729378322895,
+      "grad_norm": 1.6861886978149414,
+      "learning_rate": 9.999591260045783e-06,
+      "loss": 0.218,
+      "step": 2580
+    },
+    {
+      "epoch": 0.034215928830868034,
+      "grad_norm": 1.3947396278381348,
+      "learning_rate": 9.999534945636206e-06,
+      "loss": 0.2223,
+      "step": 2600
+    },
+    {
+      "epoch": 0.03447912828341317,
+      "grad_norm": 1.405989646911621,
+      "learning_rate": 9.999474998271246e-06,
+      "loss": 0.2153,
+      "step": 2620
+    },
+    {
+      "epoch": 0.03474232773595831,
+      "grad_norm": 1.3514268398284912,
+      "learning_rate": 9.99941141799446e-06,
+      "loss": 0.2216,
+      "step": 2640
+    },
+    {
+      "epoch": 0.03500552718850345,
+      "grad_norm": 1.7519333362579346,
+      "learning_rate": 9.999344204852053e-06,
+      "loss": 0.219,
+      "step": 2660
+    },
+    {
+      "epoch": 0.03526872664104859,
+      "grad_norm": 1.4975807666778564,
+      "learning_rate": 9.999273358892866e-06,
+      "loss": 0.2612,
+      "step": 2680
+    },
+    {
+      "epoch": 0.03553192609359373,
+      "grad_norm": 1.5119794607162476,
+      "learning_rate": 9.99919888016838e-06,
+      "loss": 0.2434,
+      "step": 2700
+    },
+    {
+      "epoch": 0.03579512554613887,
+      "grad_norm": 1.8205336332321167,
+      "learning_rate": 9.99912076873271e-06,
+      "loss": 0.2214,
+      "step": 2720
+    },
+    {
+      "epoch": 0.036058324998684,
+      "grad_norm": 1.5392286777496338,
+      "learning_rate": 9.999039024642623e-06,
+      "loss": 0.2242,
+      "step": 2740
+    },
+    {
+      "epoch": 0.03632152445122914,
+      "grad_norm": 1.9860705137252808,
+      "learning_rate": 9.99895364795752e-06,
+      "loss": 0.2238,
+      "step": 2760
+    },
+    {
+      "epoch": 0.03658472390377428,
+      "grad_norm": 1.2208589315414429,
+      "learning_rate": 9.998864638739433e-06,
+      "loss": 0.2372,
+      "step": 2780
+    },
+    {
+      "epoch": 0.036847923356319416,
+      "grad_norm": 1.0561301708221436,
+      "learning_rate": 9.998771997053047e-06,
+      "loss": 0.2543,
+      "step": 2800
+    },
+    {
+      "epoch": 0.037111122808864555,
+      "grad_norm": 1.0612584352493286,
+      "learning_rate": 9.998675722965681e-06,
+      "loss": 0.2225,
+      "step": 2820
+    },
+    {
+      "epoch": 0.037374322261409694,
+      "grad_norm": 1.527048110961914,
+      "learning_rate": 9.998575816547289e-06,
+      "loss": 0.229,
+      "step": 2840
+    },
+    {
+      "epoch": 0.03763752171395483,
+      "grad_norm": 1.3382327556610107,
+      "learning_rate": 9.998472277870474e-06,
+      "loss": 0.2269,
+      "step": 2860
+    },
+    {
+      "epoch": 0.03790072116649997,
+      "grad_norm": 1.6014606952667236,
+      "learning_rate": 9.998365107010473e-06,
+      "loss": 0.225,
+      "step": 2880
+    },
+    {
+      "epoch": 0.03816392061904511,
+      "grad_norm": 1.599460482597351,
+      "learning_rate": 9.998254304045158e-06,
+      "loss": 0.2239,
+      "step": 2900
+    },
+    {
+      "epoch": 0.03842712007159025,
+      "grad_norm": 1.5666899681091309,
+      "learning_rate": 9.998139869055048e-06,
+      "loss": 0.221,
+      "step": 2920
+    },
+    {
+      "epoch": 0.03869031952413539,
+      "grad_norm": 1.880408763885498,
+      "learning_rate": 9.9980218021233e-06,
+      "loss": 0.2322,
+      "step": 2940
+    },
+    {
+      "epoch": 0.03895351897668053,
+      "grad_norm": 1.442871332168579,
+      "learning_rate": 9.997900103335705e-06,
+      "loss": 0.2178,
+      "step": 2960
+    },
+    {
+      "epoch": 0.03921671842922567,
+      "grad_norm": 1.2973880767822266,
+      "learning_rate": 9.997774772780699e-06,
+      "loss": 0.2398,
+      "step": 2980
+    },
+    {
+      "epoch": 0.039479917881770805,
+      "grad_norm": 1.3596652746200562,
+      "learning_rate": 9.997645810549353e-06,
+      "loss": 0.2424,
+      "step": 3000
+    },
+    {
+      "epoch": 0.039479917881770805,
+      "eval_loss": 0.2598939538002014,
+      "eval_runtime": 4.1269,
+      "eval_samples_per_second": 3.877,
+      "eval_steps_per_second": 1.938,
+      "step": 3000
+    },
+    {
+      "epoch": 0.039743117334315944,
+      "grad_norm": 1.6668462753295898,
+      "learning_rate": 9.997513216735378e-06,
+      "loss": 0.2267,
+      "step": 3020
+    },
+    {
+      "epoch": 0.04000631678686108,
+      "grad_norm": 1.4699478149414062,
+      "learning_rate": 9.997376991435125e-06,
+      "loss": 0.2347,
+      "step": 3040
+    },
+    {
+      "epoch": 0.04026951623940622,
+      "grad_norm": 1.4267328977584839,
+      "learning_rate": 9.997237134747585e-06,
+      "loss": 0.2259,
+      "step": 3060
+    },
+    {
+      "epoch": 0.04053271569195136,
+      "grad_norm": 1.4341713190078735,
+      "learning_rate": 9.997093646774383e-06,
+      "loss": 0.2488,
+      "step": 3080
+    },
+    {
+      "epoch": 0.0407959151444965,
+      "grad_norm": 1.3371095657348633,
+      "learning_rate": 9.996946527619789e-06,
+      "loss": 0.2337,
+      "step": 3100
+    },
+    {
+      "epoch": 0.04105911459704164,
+      "grad_norm": 1.6772454977035522,
+      "learning_rate": 9.996795777390705e-06,
+      "loss": 0.2198,
+      "step": 3120
+    },
+    {
+      "epoch": 0.04132231404958678,
+      "grad_norm": 1.470177412033081,
+      "learning_rate": 9.996641396196678e-06,
+      "loss": 0.2317,
+      "step": 3140
+    },
+    {
+      "epoch": 0.04158551350213192,
+      "grad_norm": 1.3169437646865845,
+      "learning_rate": 9.996483384149889e-06,
+      "loss": 0.2402,
+      "step": 3160
+    },
+    {
+      "epoch": 0.041848712954677056,
+      "grad_norm": 1.2356551885604858,
+      "learning_rate": 9.99632174136516e-06,
+      "loss": 0.2242,
+      "step": 3180
+    },
+    {
+      "epoch": 0.042111912407222195,
+      "grad_norm": 1.489501714706421,
+      "learning_rate": 9.996156467959949e-06,
+      "loss": 0.2346,
+      "step": 3200
+    },
+    {
+      "epoch": 0.042375111859767334,
+      "grad_norm": 1.3025301694869995,
+      "learning_rate": 9.995987564054354e-06,
+      "loss": 0.2286,
+      "step": 3220
+    },
+    {
+      "epoch": 0.04263831131231247,
+      "grad_norm": 1.4249610900878906,
+      "learning_rate": 9.995815029771109e-06,
+      "loss": 0.2035,
+      "step": 3240
+    },
+    {
+      "epoch": 0.04290151076485761,
+      "grad_norm": 1.2967218160629272,
+      "learning_rate": 9.995638865235589e-06,
+      "loss": 0.2263,
+      "step": 3260
+    },
+    {
+      "epoch": 0.04316471021740275,
+      "grad_norm": 1.435335636138916,
+      "learning_rate": 9.995459070575806e-06,
+      "loss": 0.234,
+      "step": 3280
+    },
+    {
+      "epoch": 0.04342790966994789,
+      "grad_norm": 1.7907625436782837,
+      "learning_rate": 9.995275645922406e-06,
+      "loss": 0.2228,
+      "step": 3300
+    },
+    {
+      "epoch": 0.04369110912249303,
+      "grad_norm": 1.484876036643982,
+      "learning_rate": 9.995088591408679e-06,
+      "loss": 0.2052,
+      "step": 3320
+    },
+    {
+      "epoch": 0.04395430857503817,
+      "grad_norm": 1.4204950332641602,
+      "learning_rate": 9.994897907170552e-06,
+      "loss": 0.2279,
+      "step": 3340
+    },
+    {
+      "epoch": 0.0442175080275833,
+      "grad_norm": 1.1262263059616089,
+      "learning_rate": 9.994703593346582e-06,
+      "loss": 0.2201,
+      "step": 3360
+    },
+    {
+      "epoch": 0.04448070748012844,
+      "grad_norm": 1.1142569780349731,
+      "learning_rate": 9.994505650077972e-06,
+      "loss": 0.2307,
+      "step": 3380
+    },
+    {
+      "epoch": 0.04474390693267358,
+      "grad_norm": 1.3375637531280518,
+      "learning_rate": 9.994304077508559e-06,
+      "loss": 0.2254,
+      "step": 3400
+    },
+    {
+      "epoch": 0.045007106385218716,
+      "grad_norm": 1.7071048021316528,
+      "learning_rate": 9.994098875784818e-06,
+      "loss": 0.2251,
+      "step": 3420
+    },
+    {
+      "epoch": 0.045270305837763855,
+      "grad_norm": 1.280800461769104,
+      "learning_rate": 9.993890045055858e-06,
+      "loss": 0.2083,
+      "step": 3440
+    },
+    {
+      "epoch": 0.045533505290308994,
+      "grad_norm": 1.285821557044983,
+      "learning_rate": 9.993677585473431e-06,
+      "loss": 0.2149,
+      "step": 3460
+    },
+    {
+      "epoch": 0.04579670474285413,
+      "grad_norm": 1.6217596530914307,
+      "learning_rate": 9.993461497191921e-06,
+      "loss": 0.2193,
+      "step": 3480
+    },
+    {
+      "epoch": 0.04605990419539927,
+      "grad_norm": 1.2754729986190796,
+      "learning_rate": 9.99324178036835e-06,
+      "loss": 0.2242,
+      "step": 3500
+    },
+    {
+      "epoch": 0.04605990419539927,
+      "eval_loss": 0.25410327315330505,
+      "eval_runtime": 4.3349,
+      "eval_samples_per_second": 3.691,
+      "eval_steps_per_second": 1.846,
+      "step": 3500
+    },
+    {
+      "epoch": 0.04632310364794441,
+      "grad_norm": 1.7614775896072388,
+      "learning_rate": 9.993018435162377e-06,
+      "loss": 0.2335,
+      "step": 3520
+    },
+    {
+      "epoch": 0.04658630310048955,
+      "grad_norm": 1.1871988773345947,
+      "learning_rate": 9.992791461736298e-06,
+      "loss": 0.2411,
+      "step": 3540
+    },
+    {
+      "epoch": 0.04684950255303469,
+      "grad_norm": 1.4161990880966187,
+      "learning_rate": 9.992560860255047e-06,
+      "loss": 0.2181,
+      "step": 3560
+    },
+    {
+      "epoch": 0.04711270200557983,
+      "grad_norm": 1.5223530530929565,
+      "learning_rate": 9.99232663088619e-06,
+      "loss": 0.2394,
+      "step": 3580
+    },
+    {
+      "epoch": 0.04737590145812497,
+      "grad_norm": 1.3138693571090698,
+      "learning_rate": 9.992088773799933e-06,
+      "loss": 0.2175,
+      "step": 3600
+    },
+    {
+      "epoch": 0.047639100910670105,
+      "grad_norm": 1.5811734199523926,
+      "learning_rate": 9.991847289169118e-06,
+      "loss": 0.2078,
+      "step": 3620
+    },
+    {
+      "epoch": 0.047902300363215244,
+      "grad_norm": 1.7481496334075928,
+      "learning_rate": 9.99160217716922e-06,
+      "loss": 0.2394,
+      "step": 3640
+    },
+    {
+      "epoch": 0.04816549981576038,
+      "grad_norm": 1.6905295848846436,
+      "learning_rate": 9.99135343797835e-06,
+      "loss": 0.2197,
+      "step": 3660
+    },
+    {
+      "epoch": 0.04842869926830552,
+      "grad_norm": 1.919877052307129,
+      "learning_rate": 9.991101071777263e-06,
+      "loss": 0.2354,
+      "step": 3680
+    },
+    {
+      "epoch": 0.04869189872085066,
+      "grad_norm": 1.0950651168823242,
+      "learning_rate": 9.990845078749337e-06,
+      "loss": 0.2121,
+      "step": 3700
+    },
+    {
+      "epoch": 0.0489550981733958,
+      "grad_norm": 1.2534548044204712,
+      "learning_rate": 9.990585459080593e-06,
+      "loss": 0.2273,
+      "step": 3720
+    },
+    {
+      "epoch": 0.04921829762594094,
+      "grad_norm": 1.2291340827941895,
+      "learning_rate": 9.990322212959687e-06,
+      "loss": 0.2362,
+      "step": 3740
+    },
+    {
+      "epoch": 0.04948149707848608,
+      "grad_norm": 1.277834415435791,
+      "learning_rate": 9.990055340577909e-06,
+      "loss": 0.2142,
+      "step": 3760
+    },
+    {
+      "epoch": 0.04974469653103122,
+      "grad_norm": 1.8979097604751587,
+      "learning_rate": 9.989784842129182e-06,
+      "loss": 0.2207,
+      "step": 3780
+    },
+    {
+      "epoch": 0.050007895983576356,
+      "grad_norm": 1.5316628217697144,
+      "learning_rate": 9.989510717810069e-06,
+      "loss": 0.2194,
+      "step": 3800
+    },
+    {
+      "epoch": 0.050271095436121495,
+      "grad_norm": 1.853218674659729,
+      "learning_rate": 9.989232967819763e-06,
+      "loss": 0.2106,
+      "step": 3820
+    },
+    {
+      "epoch": 0.050534294888666634,
+      "grad_norm": 1.383473515510559,
+      "learning_rate": 9.988951592360094e-06,
+      "loss": 0.2278,
+      "step": 3840
+    },
+    {
+      "epoch": 0.05079749434121177,
+      "grad_norm": 1.4751009941101074,
+      "learning_rate": 9.988666591635527e-06,
+      "loss": 0.2211,
+      "step": 3860
+    },
+    {
+      "epoch": 0.05106069379375691,
+      "grad_norm": 1.2546067237854004,
+      "learning_rate": 9.988377965853158e-06,
+      "loss": 0.2349,
+      "step": 3880
+    },
+    {
+      "epoch": 0.05132389324630205,
+      "grad_norm": 1.477562665939331,
+      "learning_rate": 9.98808571522272e-06,
+      "loss": 0.2425,
+      "step": 3900
+    },
+    {
+      "epoch": 0.05158709269884719,
+      "grad_norm": 1.3417891263961792,
+      "learning_rate": 9.987789839956581e-06,
+      "loss": 0.2128,
+      "step": 3920
+    },
+    {
+      "epoch": 0.05185029215139233,
+      "grad_norm": 1.7570854425430298,
+      "learning_rate": 9.987490340269742e-06,
+      "loss": 0.2258,
+      "step": 3940
+    },
+    {
+      "epoch": 0.05211349160393746,
+      "grad_norm": 1.3986186981201172,
+      "learning_rate": 9.987187216379833e-06,
+      "loss": 0.2266,
+      "step": 3960
+    },
+    {
+      "epoch": 0.0523766910564826,
+      "grad_norm": 1.178642988204956,
+      "learning_rate": 9.986880468507123e-06,
+      "loss": 0.2272,
+      "step": 3980
+    },
+    {
+      "epoch": 0.05263989050902774,
+      "grad_norm": 1.6861838102340698,
+      "learning_rate": 9.986570096874514e-06,
+      "loss": 0.2333,
+      "step": 4000
+    },
+    {
+      "epoch": 0.05263989050902774,
+      "eval_loss": 0.25019052624702454,
+      "eval_runtime": 4.2429,
+      "eval_samples_per_second": 3.771,
+      "eval_steps_per_second": 1.886,
+      "step": 4000
+    },
+    {
+      "epoch": 0.05290308996157288,
+      "grad_norm": 1.8072744607925415,
+      "learning_rate": 9.986256101707543e-06,
+      "loss": 0.2301,
+      "step": 4020
+    },
+    {
+      "epoch": 0.053166289414118016,
+      "grad_norm": 1.7087039947509766,
+      "learning_rate": 9.985938483234373e-06,
+      "loss": 0.2028,
+      "step": 4040
+    },
+    {
+      "epoch": 0.053429488866663155,
+      "grad_norm": 1.1859877109527588,
+      "learning_rate": 9.985617241685806e-06,
+      "loss": 0.247,
+      "step": 4060
+    },
+    {
+      "epoch": 0.053692688319208294,
+      "grad_norm": 1.1237972974777222,
+      "learning_rate": 9.985292377295272e-06,
+      "loss": 0.2149,
+      "step": 4080
+    },
+    {
+      "epoch": 0.05395588777175343,
+      "grad_norm": 1.4817577600479126,
+      "learning_rate": 9.984963890298842e-06,
+      "loss": 0.2075,
+      "step": 4100
+    },
+    {
+      "epoch": 0.05421908722429857,
+      "grad_norm": 1.0913251638412476,
+      "learning_rate": 9.984631780935211e-06,
+      "loss": 0.2168,
+      "step": 4120
+    },
+    {
+      "epoch": 0.05448228667684371,
+      "grad_norm": 1.0020302534103394,
+      "learning_rate": 9.98429604944571e-06,
+      "loss": 0.2249,
+      "step": 4140
+    },
+    {
+      "epoch": 0.05474548612938885,
+      "grad_norm": 1.5173547267913818,
+      "learning_rate": 9.983956696074298e-06,
+      "loss": 0.2206,
+      "step": 4160
+    },
+    {
+      "epoch": 0.05500868558193399,
+      "grad_norm": 1.3773226737976074,
+      "learning_rate": 9.983613721067573e-06,
+      "loss": 0.2125,
+      "step": 4180
+    },
+    {
+      "epoch": 0.05527188503447913,
+      "grad_norm": 1.38734769821167,
+      "learning_rate": 9.983267124674759e-06,
+      "loss": 0.2328,
+      "step": 4200
+    },
+    {
+      "epoch": 0.05553508448702427,
+      "grad_norm": 1.764356255531311,
+      "learning_rate": 9.982916907147713e-06,
+      "loss": 0.2224,
+      "step": 4220
+    },
+    {
+      "epoch": 0.055798283939569406,
+      "grad_norm": 1.3347185850143433,
+      "learning_rate": 9.982563068740925e-06,
+      "loss": 0.2131,
+      "step": 4240
+    },
+    {
+      "epoch": 0.056061483392114544,
+      "grad_norm": 1.8289685249328613,
+      "learning_rate": 9.982205609711515e-06,
+      "loss": 0.2314,
+      "step": 4260
+    },
+    {
+      "epoch": 0.05632468284465968,
+      "grad_norm": 1.5655499696731567,
+      "learning_rate": 9.981844530319231e-06,
+      "loss": 0.2289,
+      "step": 4280
+    },
+    {
+      "epoch": 0.05658788229720482,
+      "grad_norm": 1.2388875484466553,
+      "learning_rate": 9.981479830826454e-06,
+      "loss": 0.2142,
+      "step": 4300
+    },
+    {
+      "epoch": 0.05685108174974996,
+      "grad_norm": 1.5092289447784424,
+      "learning_rate": 9.981111511498201e-06,
+      "loss": 0.2272,
+      "step": 4320
+    },
+    {
+      "epoch": 0.0571142812022951,
+      "grad_norm": 1.590738296508789,
+      "learning_rate": 9.980739572602112e-06,
+      "loss": 0.2281,
+      "step": 4340
+    },
+    {
+      "epoch": 0.05737748065484024,
+      "grad_norm": 1.8792939186096191,
+      "learning_rate": 9.980364014408457e-06,
+      "loss": 0.2013,
+      "step": 4360
+    },
+    {
+      "epoch": 0.05764068010738538,
+      "grad_norm": 1.8997962474822998,
+      "learning_rate": 9.979984837190142e-06,
+      "loss": 0.2154,
+      "step": 4380
+    },
+    {
+      "epoch": 0.05790387955993052,
+      "grad_norm": 1.481593370437622,
+      "learning_rate": 9.979602041222699e-06,
+      "loss": 0.2318,
+      "step": 4400
+    },
+    {
+      "epoch": 0.058167079012475656,
+      "grad_norm": 1.6572890281677246,
+      "learning_rate": 9.979215626784288e-06,
+      "loss": 0.2022,
+      "step": 4420
+    },
+    {
+      "epoch": 0.058430278465020795,
+      "grad_norm": 1.2502589225769043,
+      "learning_rate": 9.978825594155701e-06,
+      "loss": 0.2149,
+      "step": 4440
+    },
+    {
+      "epoch": 0.058693477917565934,
+      "grad_norm": 1.8111909627914429,
+      "learning_rate": 9.97843194362036e-06,
+      "loss": 0.2051,
+      "step": 4460
+    },
+    {
+      "epoch": 0.05895667737011107,
+      "grad_norm": 1.3169831037521362,
+      "learning_rate": 9.978034675464314e-06,
+      "loss": 0.2111,
+      "step": 4480
+    },
+    {
+      "epoch": 0.05921987682265621,
+      "grad_norm": 1.7849903106689453,
+      "learning_rate": 9.977633789976242e-06,
+      "loss": 0.2162,
+      "step": 4500
+    },
+    {
+      "epoch": 0.05921987682265621,
+      "eval_loss": 0.258551687002182,
+      "eval_runtime": 4.4268,
+      "eval_samples_per_second": 3.614,
+      "eval_steps_per_second": 1.807,
+      "step": 4500
+    },
+    {
+      "epoch": 0.05948307627520135,
+      "grad_norm": 1.8503687381744385,
+      "learning_rate": 9.977229287447447e-06,
+      "loss": 0.2173,
+      "step": 4520
+    },
+    {
+      "epoch": 0.05974627572774649,
+      "grad_norm": 1.5168489217758179,
+      "learning_rate": 9.97682116817187e-06,
+      "loss": 0.2282,
+      "step": 4540
+    },
+    {
+      "epoch": 0.06000947518029162,
+      "grad_norm": 1.632920503616333,
+      "learning_rate": 9.97640943244607e-06,
+      "loss": 0.2333,
+      "step": 4560
+    },
+    {
+      "epoch": 0.06027267463283676,
+      "grad_norm": 1.28827702999115,
+      "learning_rate": 9.975994080569241e-06,
+      "loss": 0.2186,
+      "step": 4580
+    },
+    {
+      "epoch": 0.0605358740853819,
+      "grad_norm": 1.9723427295684814,
+      "learning_rate": 9.975575112843201e-06,
+      "loss": 0.2121,
+      "step": 4600
+    },
+    {
+      "epoch": 0.06079907353792704,
+      "grad_norm": 1.6105324029922485,
+      "learning_rate": 9.975152529572396e-06,
+      "loss": 0.2043,
+      "step": 4620
+    },
+    {
+      "epoch": 0.06106227299047218,
+      "grad_norm": 1.242079734802246,
+      "learning_rate": 9.974726331063902e-06,
+      "loss": 0.2058,
+      "step": 4640
+    },
+    {
+      "epoch": 0.061325472443017316,
+      "grad_norm": 1.250302791595459,
+      "learning_rate": 9.974296517627417e-06,
+      "loss": 0.2241,
+      "step": 4660
+    },
+    {
+      "epoch": 0.061588671895562455,
+      "grad_norm": 1.492856502532959,
+      "learning_rate": 9.973863089575269e-06,
+      "loss": 0.2121,
+      "step": 4680
+    },
+    {
+      "epoch": 0.061851871348107594,
+      "grad_norm": 1.7203097343444824,
+      "learning_rate": 9.973426047222415e-06,
+      "loss": 0.2382,
+      "step": 4700
+    },
+    {
+      "epoch": 0.06211507080065273,
+      "grad_norm": 1.4408618211746216,
+      "learning_rate": 9.972985390886432e-06,
+      "loss": 0.2279,
+      "step": 4720
+    },
+    {
+      "epoch": 0.06237827025319787,
+      "grad_norm": 0.8927859663963318,
+      "learning_rate": 9.97254112088753e-06,
+      "loss": 0.2337,
+      "step": 4740
+    },
+    {
+      "epoch": 0.06264146970574301,
+      "grad_norm": 1.5471599102020264,
+      "learning_rate": 9.972093237548538e-06,
+      "loss": 0.2022,
+      "step": 4760
+    },
+    {
+      "epoch": 0.06290466915828816,
+      "grad_norm": 1.4374244213104248,
+      "learning_rate": 9.971641741194918e-06,
+      "loss": 0.244,
+      "step": 4780
+    },
+    {
+      "epoch": 0.06316786861083329,
+      "grad_norm": 1.2770164012908936,
+      "learning_rate": 9.971186632154753e-06,
+      "loss": 0.2235,
+      "step": 4800
+    },
+    {
+      "epoch": 0.06343106806337843,
+      "grad_norm": 1.2279208898544312,
+      "learning_rate": 9.97072791075875e-06,
+      "loss": 0.2407,
+      "step": 4820
+    },
+    {
+      "epoch": 0.06369426751592357,
+      "grad_norm": 1.5687545537948608,
+      "learning_rate": 9.970265577340243e-06,
+      "loss": 0.2118,
+      "step": 4840
+    },
+    {
+      "epoch": 0.06395746696846871,
+      "grad_norm": 1.8475176095962524,
+      "learning_rate": 9.969799632235193e-06,
+      "loss": 0.2282,
+      "step": 4860
+    },
+    {
+      "epoch": 0.06422066642101384,
+      "grad_norm": 1.502355933189392,
+      "learning_rate": 9.969330075782183e-06,
+      "loss": 0.2209,
+      "step": 4880
+    },
+    {
+      "epoch": 0.06448386587355898,
+      "grad_norm": 1.8428897857666016,
+      "learning_rate": 9.968856908322415e-06,
+      "loss": 0.2311,
+      "step": 4900
+    },
+    {
+      "epoch": 0.06474706532610412,
+      "grad_norm": 1.460692048072815,
+      "learning_rate": 9.968380130199726e-06,
+      "loss": 0.2096,
+      "step": 4920
+    },
+    {
+      "epoch": 0.06501026477864925,
+      "grad_norm": 1.203684687614441,
+      "learning_rate": 9.967899741760571e-06,
+      "loss": 0.2298,
+      "step": 4940
+    },
+    {
+      "epoch": 0.0652734642311944,
+      "grad_norm": 1.6590291261672974,
+      "learning_rate": 9.967415743354024e-06,
+      "loss": 0.2277,
+      "step": 4960
+    },
+    {
+      "epoch": 0.06553666368373953,
+      "grad_norm": 1.221056342124939,
+      "learning_rate": 9.96692813533179e-06,
+      "loss": 0.2236,
+      "step": 4980
+    },
+    {
+      "epoch": 0.06579986313628468,
+      "grad_norm": 1.269662857055664,
+      "learning_rate": 9.966436918048193e-06,
+      "loss": 0.2192,
+      "step": 5000
+    },
+    {
+      "epoch": 0.06579986313628468,
+      "eval_loss": 0.2526422142982483,
+      "eval_runtime": 4.4276,
+      "eval_samples_per_second": 3.614,
+      "eval_steps_per_second": 1.807,
+      "step": 5000
+    },
+    {
+      "epoch": 0.06606306258882981,
+      "grad_norm": 1.1057158708572388,
+      "learning_rate": 9.96594209186018e-06,
+      "loss": 0.1944,
+      "step": 5020
+    },
+    {
+      "epoch": 0.06632626204137496,
+      "grad_norm": 1.9506696462631226,
+      "learning_rate": 9.965443657127319e-06,
+      "loss": 0.2269,
+      "step": 5040
+    },
+    {
+      "epoch": 0.06658946149392009,
+      "grad_norm": 1.3153961896896362,
+      "learning_rate": 9.964941614211805e-06,
+      "loss": 0.2047,
+      "step": 5060
+    },
+    {
+      "epoch": 0.06685266094646523,
+      "grad_norm": 1.2307437658309937,
+      "learning_rate": 9.96443596347845e-06,
+      "loss": 0.2221,
+      "step": 5080
+    },
+    {
+      "epoch": 0.06711586039901037,
+      "grad_norm": 1.3759076595306396,
+      "learning_rate": 9.963926705294691e-06,
+      "loss": 0.2098,
+      "step": 5100
+    },
+    {
+      "epoch": 0.06737905985155551,
+      "grad_norm": 1.584091305732727,
+      "learning_rate": 9.963413840030584e-06,
+      "loss": 0.2135,
+      "step": 5120
+    },
+    {
+      "epoch": 0.06764225930410064,
+      "grad_norm": 1.4988020658493042,
+      "learning_rate": 9.962897368058807e-06,
+      "loss": 0.2262,
+      "step": 5140
+    },
+    {
+      "epoch": 0.06790545875664579,
+      "grad_norm": 1.7893013954162598,
+      "learning_rate": 9.962377289754657e-06,
+      "loss": 0.2095,
+      "step": 5160
+    },
+    {
+      "epoch": 0.06816865820919092,
+      "grad_norm": 1.5329294204711914,
+      "learning_rate": 9.961853605496056e-06,
+      "loss": 0.2231,
+      "step": 5180
+    },
+    {
+      "epoch": 0.06843185766173607,
+      "grad_norm": 1.1528375148773193,
+      "learning_rate": 9.961326315663542e-06,
+      "loss": 0.2391,
+      "step": 5200
+    },
+    {
+      "epoch": 0.0686950571142812,
+      "grad_norm": 1.1562128067016602,
+      "learning_rate": 9.960795420640278e-06,
+      "loss": 0.2227,
+      "step": 5220
+    },
+    {
+      "epoch": 0.06895825656682635,
+      "grad_norm": 1.373683214187622,
+      "learning_rate": 9.960260920812038e-06,
+      "loss": 0.2169,
+      "step": 5240
+    },
+    {
+      "epoch": 0.06922145601937148,
+      "grad_norm": 1.3141005039215088,
+      "learning_rate": 9.959722816567226e-06,
+      "loss": 0.2197,
+      "step": 5260
+    },
+    {
+      "epoch": 0.06948465547191662,
+      "grad_norm": 1.3508551120758057,
+      "learning_rate": 9.959181108296856e-06,
+      "loss": 0.2198,
+      "step": 5280
+    },
+    {
+      "epoch": 0.06974785492446176,
+      "grad_norm": 1.4236295223236084,
+      "learning_rate": 9.958635796394569e-06,
+      "loss": 0.213,
+      "step": 5300
+    },
+    {
+      "epoch": 0.0700110543770069,
+      "grad_norm": 1.4767459630966187,
+      "learning_rate": 9.958086881256616e-06,
+      "loss": 0.2143,
+      "step": 5320
+    },
+    {
+      "epoch": 0.07027425382955203,
+      "grad_norm": 1.598724126815796,
+      "learning_rate": 9.957534363281876e-06,
+      "loss": 0.2201,
+      "step": 5340
+    },
+    {
+      "epoch": 0.07053745328209718,
+      "grad_norm": 1.9216469526290894,
+      "learning_rate": 9.956978242871838e-06,
+      "loss": 0.2075,
+      "step": 5360
+    },
+    {
+      "epoch": 0.07080065273464231,
+      "grad_norm": 1.6237411499023438,
+      "learning_rate": 9.956418520430613e-06,
+      "loss": 0.1957,
+      "step": 5380
+    },
+    {
+      "epoch": 0.07106385218718746,
+      "grad_norm": 1.6576958894729614,
+      "learning_rate": 9.955855196364927e-06,
+      "loss": 0.2154,
+      "step": 5400
+    },
+    {
+      "epoch": 0.07132705163973259,
+      "grad_norm": 1.5656834840774536,
+      "learning_rate": 9.955288271084124e-06,
+      "loss": 0.2082,
+      "step": 5420
+    },
+    {
+      "epoch": 0.07159025109227773,
+      "grad_norm": 2.2170300483703613,
+      "learning_rate": 9.954717745000166e-06,
+      "loss": 0.2275,
+      "step": 5440
+    },
+    {
+      "epoch": 0.07185345054482287,
+      "grad_norm": 1.2994457483291626,
+      "learning_rate": 9.954143618527632e-06,
+      "loss": 0.2097,
+      "step": 5460
+    },
+    {
+      "epoch": 0.072116649997368,
+      "grad_norm": 1.500351071357727,
+      "learning_rate": 9.953565892083713e-06,
+      "loss": 0.2158,
+      "step": 5480
+    },
+    {
+      "epoch": 0.07237984944991314,
+      "grad_norm": 1.0603489875793457,
+      "learning_rate": 9.952984566088221e-06,
+      "loss": 0.2329,
+      "step": 5500
+    },
+    {
+      "epoch": 0.07237984944991314,
+      "eval_loss": 0.24393533170223236,
+      "eval_runtime": 4.248,
+      "eval_samples_per_second": 3.766,
+      "eval_steps_per_second": 1.883,
+      "step": 5500
+    },
+    {
+      "epoch": 0.07264304890245828,
+      "grad_norm": 1.3520938158035278,
+      "learning_rate": 9.952399640963582e-06,
+      "loss": 0.2088,
+      "step": 5520
+    },
+    {
+      "epoch": 0.07290624835500342,
+      "grad_norm": 1.4677773714065552,
+      "learning_rate": 9.951811117134835e-06,
+      "loss": 0.2138,
+      "step": 5540
+    },
+    {
+      "epoch": 0.07316944780754855,
+      "grad_norm": 1.5744502544403076,
+      "learning_rate": 9.951218995029637e-06,
+      "loss": 0.2155,
+      "step": 5560
+    },
+    {
+      "epoch": 0.0734326472600937,
+      "grad_norm": 1.258615255355835,
+      "learning_rate": 9.95062327507826e-06,
+      "loss": 0.2163,
+      "step": 5580
+    },
+    {
+      "epoch": 0.07369584671263883,
+      "grad_norm": 0.9961044192314148,
+      "learning_rate": 9.950023957713586e-06,
+      "loss": 0.22,
+      "step": 5600
+    },
+    {
+      "epoch": 0.07395904616518398,
+      "grad_norm": 1.1791836023330688,
+      "learning_rate": 9.949421043371115e-06,
+      "loss": 0.2258,
+      "step": 5620
+    },
+    {
+      "epoch": 0.07422224561772911,
+      "grad_norm": 1.551808476448059,
+      "learning_rate": 9.948814532488963e-06,
+      "loss": 0.2147,
+      "step": 5640
+    },
+    {
+      "epoch": 0.07448544507027426,
+      "grad_norm": 1.8343244791030884,
+      "learning_rate": 9.948204425507853e-06,
+      "loss": 0.2064,
+      "step": 5660
+    },
+    {
+      "epoch": 0.07474864452281939,
+      "grad_norm": 1.501928448677063,
+      "learning_rate": 9.947590722871126e-06,
+      "loss": 0.2246,
+      "step": 5680
+    },
+    {
+      "epoch": 0.07501184397536453,
+      "grad_norm": 1.3524726629257202,
+      "learning_rate": 9.946973425024734e-06,
+      "loss": 0.2257,
+      "step": 5700
+    },
+    {
+      "epoch": 0.07527504342790967,
+      "grad_norm": 1.4547368288040161,
+      "learning_rate": 9.946352532417242e-06,
+      "loss": 0.2082,
+      "step": 5720
+    },
+    {
+      "epoch": 0.07553824288045481,
+      "grad_norm": 1.3905247449874878,
+      "learning_rate": 9.945728045499827e-06,
+      "loss": 0.2103,
+      "step": 5740
+    },
+    {
+      "epoch": 0.07580144233299994,
+      "grad_norm": 1.6494786739349365,
+      "learning_rate": 9.945099964726278e-06,
+      "loss": 0.2231,
+      "step": 5760
+    },
+    {
+      "epoch": 0.07606464178554509,
+      "grad_norm": 1.6413875818252563,
+      "learning_rate": 9.944468290552992e-06,
+      "loss": 0.2433,
+      "step": 5780
+    },
+    {
+      "epoch": 0.07632784123809022,
+      "grad_norm": 1.3650838136672974,
+      "learning_rate": 9.943833023438987e-06,
+      "loss": 0.203,
+      "step": 5800
+    },
+    {
+      "epoch": 0.07659104069063537,
+      "grad_norm": 1.248500943183899,
+      "learning_rate": 9.943194163845881e-06,
+      "loss": 0.2268,
+      "step": 5820
+    },
+    {
+      "epoch": 0.0768542401431805,
+      "grad_norm": 1.6765228509902954,
+      "learning_rate": 9.942551712237907e-06,
+      "loss": 0.2217,
+      "step": 5840
+    },
+    {
+      "epoch": 0.07711743959572565,
+      "grad_norm": 1.2180379629135132,
+      "learning_rate": 9.941905669081909e-06,
+      "loss": 0.2116,
+      "step": 5860
+    },
+    {
+      "epoch": 0.07738063904827078,
+      "grad_norm": 1.7382229566574097,
+      "learning_rate": 9.941256034847339e-06,
+      "loss": 0.2082,
+      "step": 5880
+    },
+    {
+      "epoch": 0.07764383850081592,
+      "grad_norm": 1.522442102432251,
+      "learning_rate": 9.940602810006262e-06,
+      "loss": 0.2169,
+      "step": 5900
+    },
+    {
+      "epoch": 0.07790703795336106,
+      "grad_norm": 1.4671295881271362,
+      "learning_rate": 9.939945995033346e-06,
+      "loss": 0.2326,
+      "step": 5920
+    },
+    {
+      "epoch": 0.0781702374059062,
+      "grad_norm": 1.1275602579116821,
+      "learning_rate": 9.939285590405873e-06,
+      "loss": 0.2281,
+      "step": 5940
+    },
+    {
+      "epoch": 0.07843343685845133,
+      "grad_norm": 1.3872261047363281,
+      "learning_rate": 9.938621596603732e-06,
+      "loss": 0.216,
+      "step": 5960
+    },
+    {
+      "epoch": 0.07869663631099648,
+      "grad_norm": 1.3105968236923218,
+      "learning_rate": 9.937954014109419e-06,
+      "loss": 0.22,
+      "step": 5980
+    },
+    {
+      "epoch": 0.07895983576354161,
+      "grad_norm": 1.2344915866851807,
+      "learning_rate": 9.937282843408039e-06,
+      "loss": 0.2133,
+      "step": 6000
+    },
+    {
+      "epoch": 0.07895983576354161,
+      "eval_loss": 0.24625569581985474,
+      "eval_runtime": 4.2742,
+      "eval_samples_per_second": 3.743,
+      "eval_steps_per_second": 1.872,
+      "step": 6000
+    },
+    {
+      "epoch": 0.07922303521608676,
+      "grad_norm": 1.9293681383132935,
+      "learning_rate": 9.936608084987302e-06,
+      "loss": 0.2249,
+      "step": 6020
+    },
+    {
+      "epoch": 0.07948623466863189,
+      "grad_norm": 1.61739182472229,
+      "learning_rate": 9.935929739337531e-06,
+      "loss": 0.2137,
+      "step": 6040
+    },
+    {
+      "epoch": 0.07974943412117703,
+      "grad_norm": 1.5952469110488892,
+      "learning_rate": 9.93524780695165e-06,
+      "loss": 0.2053,
+      "step": 6060
+    },
+    {
+      "epoch": 0.08001263357372217,
+      "grad_norm": 1.2378883361816406,
+      "learning_rate": 9.93456228832519e-06,
+      "loss": 0.2067,
+      "step": 6080
+    },
+    {
+      "epoch": 0.0802758330262673,
+      "grad_norm": 1.6271859407424927,
+      "learning_rate": 9.93387318395629e-06,
+      "loss": 0.2061,
+      "step": 6100
+    },
+    {
+      "epoch": 0.08053903247881244,
+      "grad_norm": 1.7112281322479248,
+      "learning_rate": 9.933180494345693e-06,
+      "loss": 0.2174,
+      "step": 6120
+    },
+    {
+      "epoch": 0.08080223193135758,
+      "grad_norm": 1.7119606733322144,
+      "learning_rate": 9.93248421999675e-06,
+      "loss": 0.2269,
+      "step": 6140
+    },
+    {
+      "epoch": 0.08106543138390272,
+      "grad_norm": 1.3291398286819458,
+      "learning_rate": 9.931784361415413e-06,
+      "loss": 0.22,
+      "step": 6160
+    },
+    {
+      "epoch": 0.08132863083644785,
+      "grad_norm": 1.8670836687088013,
+      "learning_rate": 9.93108091911024e-06,
+      "loss": 0.2153,
+      "step": 6180
+    },
+    {
+      "epoch": 0.081591830288993,
+      "grad_norm": 1.5941790342330933,
+      "learning_rate": 9.930373893592392e-06,
+      "loss": 0.2124,
+      "step": 6200
+    },
+    {
+      "epoch": 0.08185502974153813,
+      "grad_norm": 1.3274775743484497,
+      "learning_rate": 9.929663285375638e-06,
+      "loss": 0.2269,
+      "step": 6220
+    },
+    {
+      "epoch": 0.08211822919408328,
+      "grad_norm": 1.6661081314086914,
+      "learning_rate": 9.928949094976347e-06,
+      "loss": 0.2197,
+      "step": 6240
+    },
+    {
+      "epoch": 0.08238142864662841,
+      "grad_norm": 1.743621587753296,
+      "learning_rate": 9.928231322913494e-06,
+      "loss": 0.2069,
+      "step": 6260
+    },
+    {
+      "epoch": 0.08264462809917356,
+      "grad_norm": 1.2716785669326782,
+      "learning_rate": 9.92750996970865e-06,
+      "loss": 0.2125,
+      "step": 6280
+    },
+    {
+      "epoch": 0.08290782755171869,
+      "grad_norm": 1.5045121908187866,
+      "learning_rate": 9.926785035885993e-06,
+      "loss": 0.2109,
+      "step": 6300
+    },
+    {
+      "epoch": 0.08317102700426383,
+      "grad_norm": 1.5263692140579224,
+      "learning_rate": 9.926056521972305e-06,
+      "loss": 0.2141,
+      "step": 6320
+    },
+    {
+      "epoch": 0.08343422645680897,
+      "grad_norm": 1.1671802997589111,
+      "learning_rate": 9.925324428496965e-06,
+      "loss": 0.203,
+      "step": 6340
+    },
+    {
+      "epoch": 0.08369742590935411,
+      "grad_norm": 1.8479175567626953,
+      "learning_rate": 9.924588755991958e-06,
+      "loss": 0.2177,
+      "step": 6360
+    },
+    {
+      "epoch": 0.08396062536189924,
+      "grad_norm": 1.419536828994751,
+      "learning_rate": 9.923849504991862e-06,
+      "loss": 0.212,
+      "step": 6380
+    },
+    {
+      "epoch": 0.08422382481444439,
+      "grad_norm": 1.902773141860962,
+      "learning_rate": 9.923106676033863e-06,
+      "loss": 0.2103,
+      "step": 6400
+    },
+    {
+      "epoch": 0.08448702426698952,
+      "grad_norm": 1.438146710395813,
+      "learning_rate": 9.922360269657745e-06,
+      "loss": 0.2229,
+      "step": 6420
+    },
+    {
+      "epoch": 0.08475022371953467,
+      "grad_norm": 1.320561170578003,
+      "learning_rate": 9.921610286405889e-06,
+      "loss": 0.219,
+      "step": 6440
+    },
+    {
+      "epoch": 0.0850134231720798,
+      "grad_norm": 1.8045339584350586,
+      "learning_rate": 9.920856726823276e-06,
+      "loss": 0.2086,
+      "step": 6460
+    },
+    {
+      "epoch": 0.08527662262462495,
+      "grad_norm": 1.1669182777404785,
+      "learning_rate": 9.920099591457488e-06,
+      "loss": 0.229,
+      "step": 6480
+    },
+    {
+      "epoch": 0.08553982207717008,
+      "grad_norm": 1.3616862297058105,
+      "learning_rate": 9.919338880858707e-06,
+      "loss": 0.2164,
+      "step": 6500
+    },
+    {
+      "epoch": 0.08553982207717008,
+      "eval_loss": 0.2475052773952484,
+      "eval_runtime": 4.3564,
+      "eval_samples_per_second": 3.673,
+      "eval_steps_per_second": 1.836,
+      "step": 6500
+    },
+    {
+      "epoch": 0.08580302152971522,
+      "grad_norm": 1.454439401626587,
+      "learning_rate": 9.918574595579703e-06,
+      "loss": 0.2079,
+      "step": 6520
+    },
+    {
+      "epoch": 0.08606622098226036,
+      "grad_norm": 1.8043402433395386,
+      "learning_rate": 9.917806736175857e-06,
+      "loss": 0.2339,
+      "step": 6540
+    },
+    {
+      "epoch": 0.0863294204348055,
+      "grad_norm": 1.5793465375900269,
+      "learning_rate": 9.917035303205136e-06,
+      "loss": 0.2109,
+      "step": 6560
+    },
+    {
+      "epoch": 0.08659261988735063,
+      "grad_norm": 1.7203484773635864,
+      "learning_rate": 9.91626029722811e-06,
+      "loss": 0.2176,
+      "step": 6580
+    },
+    {
+      "epoch": 0.08685581933989578,
+      "grad_norm": 2.077418565750122,
+      "learning_rate": 9.915481718807946e-06,
+      "loss": 0.2156,
+      "step": 6600
+    },
+    {
+      "epoch": 0.08711901879244091,
+      "grad_norm": 1.3149522542953491,
+      "learning_rate": 9.914699568510404e-06,
+      "loss": 0.2204,
+      "step": 6620
+    },
+    {
+      "epoch": 0.08738221824498606,
+      "grad_norm": 1.3341245651245117,
+      "learning_rate": 9.913913846903836e-06,
+      "loss": 0.2252,
+      "step": 6640
+    },
+    {
+      "epoch": 0.08764541769753119,
+      "grad_norm": 2.3080759048461914,
+      "learning_rate": 9.913124554559198e-06,
+      "loss": 0.2094,
+      "step": 6660
+    },
+    {
+      "epoch": 0.08790861715007633,
+      "grad_norm": 1.7658042907714844,
+      "learning_rate": 9.912331692050036e-06,
+      "loss": 0.1995,
+      "step": 6680
+    },
+    {
+      "epoch": 0.08817181660262147,
+      "grad_norm": 1.4984004497528076,
+      "learning_rate": 9.91153525995249e-06,
+      "loss": 0.2101,
+      "step": 6700
+    },
+    {
+      "epoch": 0.0884350160551666,
+      "grad_norm": 1.731812834739685,
+      "learning_rate": 9.910735258845292e-06,
+      "loss": 0.208,
+      "step": 6720
+    },
+    {
+      "epoch": 0.08869821550771174,
+      "grad_norm": 1.4385346174240112,
+      "learning_rate": 9.90993168930977e-06,
+      "loss": 0.1961,
+      "step": 6740
+    },
+    {
+      "epoch": 0.08896141496025688,
+      "grad_norm": 1.7041187286376953,
+      "learning_rate": 9.909124551929849e-06,
+      "loss": 0.2264,
+      "step": 6760
+    },
+    {
+      "epoch": 0.08922461441280202,
+      "grad_norm": 1.437453031539917,
+      "learning_rate": 9.908313847292036e-06,
+      "loss": 0.2107,
+      "step": 6780
+    },
+    {
+      "epoch": 0.08948781386534715,
+      "grad_norm": 1.7059805393218994,
+      "learning_rate": 9.907499575985443e-06,
+      "loss": 0.2208,
+      "step": 6800
+    },
+    {
+      "epoch": 0.0897510133178923,
+      "grad_norm": 1.9154585599899292,
+      "learning_rate": 9.906681738601762e-06,
+      "loss": 0.2093,
+      "step": 6820
+    },
+    {
+      "epoch": 0.09001421277043743,
+      "grad_norm": 1.1896253824234009,
+      "learning_rate": 9.905860335735287e-06,
+      "loss": 0.2383,
+      "step": 6840
+    },
+    {
+      "epoch": 0.09027741222298258,
+      "grad_norm": 1.363857626914978,
+      "learning_rate": 9.905035367982891e-06,
+      "loss": 0.2247,
+      "step": 6860
+    },
+    {
+      "epoch": 0.09054061167552771,
+      "grad_norm": 1.8386965990066528,
+      "learning_rate": 9.90420683594405e-06,
+      "loss": 0.2081,
+      "step": 6880
+    },
+    {
+      "epoch": 0.09080381112807286,
+      "grad_norm": 1.338268518447876,
+      "learning_rate": 9.903374740220818e-06,
+      "loss": 0.2033,
+      "step": 6900
+    },
+    {
+      "epoch": 0.09106701058061799,
+      "grad_norm": 1.6110661029815674,
+      "learning_rate": 9.902539081417851e-06,
+      "loss": 0.2186,
+      "step": 6920
+    },
+    {
+      "epoch": 0.09133021003316313,
+      "grad_norm": 1.3898533582687378,
+      "learning_rate": 9.901699860142384e-06,
+      "loss": 0.2172,
+      "step": 6940
+    },
+    {
+      "epoch": 0.09159340948570827,
+      "grad_norm": 1.9416710138320923,
+      "learning_rate": 9.900857077004244e-06,
+      "loss": 0.2121,
+      "step": 6960
+    },
+    {
+      "epoch": 0.09185660893825341,
+      "grad_norm": 1.195512294769287,
+      "learning_rate": 9.900010732615848e-06,
+      "loss": 0.2274,
+      "step": 6980
+    },
+    {
+      "epoch": 0.09211980839079854,
+      "grad_norm": 1.20065176486969,
+      "learning_rate": 9.8991608275922e-06,
+      "loss": 0.2292,
+      "step": 7000
+    },
+    {
+      "epoch": 0.09211980839079854,
+      "eval_loss": 0.2513369917869568,
+      "eval_runtime": 3.6345,
+      "eval_samples_per_second": 4.402,
+      "eval_steps_per_second": 2.201,
+      "step": 7000
+    },
+    {
+      "epoch": 0.09238300784334369,
+      "grad_norm": 1.3137370347976685,
+      "learning_rate": 9.89830736255089e-06,
+      "loss": 0.2154,
+      "step": 7020
+    },
+    {
+      "epoch": 0.09264620729588882,
+      "grad_norm": 1.1988691091537476,
+      "learning_rate": 9.897450338112093e-06,
+      "loss": 0.2252,
+      "step": 7040
+    },
+    {
+      "epoch": 0.09290940674843397,
+      "grad_norm": 1.3026483058929443,
+      "learning_rate": 9.89658975489858e-06,
+      "loss": 0.2234,
+      "step": 7060
+    },
+    {
+      "epoch": 0.0931726062009791,
+      "grad_norm": 1.7757319211959839,
+      "learning_rate": 9.895725613535695e-06,
+      "loss": 0.199,
+      "step": 7080
+    },
+    {
+      "epoch": 0.09343580565352425,
+      "grad_norm": 1.4556313753128052,
+      "learning_rate": 9.894857914651377e-06,
+      "loss": 0.2077,
+      "step": 7100
+    },
+    {
+      "epoch": 0.09369900510606938,
+      "grad_norm": 1.395593523979187,
+      "learning_rate": 9.893986658876144e-06,
+      "loss": 0.2127,
+      "step": 7120
+    },
+    {
+      "epoch": 0.09396220455861452,
+      "grad_norm": 1.0599098205566406,
+      "learning_rate": 9.893111846843107e-06,
+      "loss": 0.2067,
+      "step": 7140
+    },
+    {
+      "epoch": 0.09422540401115966,
+      "grad_norm": 1.5176869630813599,
+      "learning_rate": 9.89223347918795e-06,
+      "loss": 0.2286,
+      "step": 7160
+    },
+    {
+      "epoch": 0.0944886034637048,
+      "grad_norm": 1.3846049308776855,
+      "learning_rate": 9.89135155654895e-06,
+      "loss": 0.207,
+      "step": 7180
+    },
+    {
+      "epoch": 0.09475180291624993,
+      "grad_norm": 1.7676697969436646,
+      "learning_rate": 9.89046607956696e-06,
+      "loss": 0.2073,
+      "step": 7200
+    },
+    {
+      "epoch": 0.09501500236879508,
+      "grad_norm": 1.7330838441848755,
+      "learning_rate": 9.889577048885424e-06,
+      "loss": 0.2131,
+      "step": 7220
+    },
+    {
+      "epoch": 0.09527820182134021,
+      "grad_norm": 1.748367190361023,
+      "learning_rate": 9.888684465150362e-06,
+      "loss": 0.2238,
+      "step": 7240
+    },
+    {
+      "epoch": 0.09554140127388536,
+      "grad_norm": 2.1994078159332275,
+      "learning_rate": 9.887788329010378e-06,
+      "loss": 0.2026,
+      "step": 7260
+    },
+    {
+      "epoch": 0.09580460072643049,
+      "grad_norm": 1.5093474388122559,
+      "learning_rate": 9.886888641116655e-06,
+      "loss": 0.2397,
+      "step": 7280
+    },
+    {
+      "epoch": 0.09606780017897562,
+      "grad_norm": 1.1632699966430664,
+      "learning_rate": 9.885985402122961e-06,
+      "loss": 0.2296,
+      "step": 7300
+    },
+    {
+      "epoch": 0.09633099963152077,
+      "grad_norm": 1.4018467664718628,
+      "learning_rate": 9.885078612685642e-06,
+      "loss": 0.2106,
+      "step": 7320
+    },
+    {
+      "epoch": 0.0965941990840659,
+      "grad_norm": 1.3585357666015625,
+      "learning_rate": 9.884168273463625e-06,
+      "loss": 0.197,
+      "step": 7340
+    },
+    {
+      "epoch": 0.09685739853661104,
+      "grad_norm": 1.5368624925613403,
+      "learning_rate": 9.883254385118415e-06,
+      "loss": 0.2097,
+      "step": 7360
+    },
+    {
+      "epoch": 0.09712059798915618,
+      "grad_norm": 1.1643812656402588,
+      "learning_rate": 9.882336948314097e-06,
+      "loss": 0.2075,
+      "step": 7380
+    },
+    {
+      "epoch": 0.09738379744170132,
+      "grad_norm": 1.3231298923492432,
+      "learning_rate": 9.881415963717336e-06,
+      "loss": 0.2192,
+      "step": 7400
+    },
+    {
+      "epoch": 0.09764699689424645,
+      "grad_norm": 1.6496835947036743,
+      "learning_rate": 9.880491431997372e-06,
+      "loss": 0.2047,
+      "step": 7420
+    },
+    {
+      "epoch": 0.0979101963467916,
+      "grad_norm": 1.2273615598678589,
+      "learning_rate": 9.879563353826023e-06,
+      "loss": 0.2109,
+      "step": 7440
+    },
+    {
+      "epoch": 0.09817339579933673,
+      "grad_norm": 1.558263897895813,
+      "learning_rate": 9.878631729877686e-06,
+      "loss": 0.2023,
+      "step": 7460
+    },
+    {
+      "epoch": 0.09843659525188188,
+      "grad_norm": 1.6220104694366455,
+      "learning_rate": 9.877696560829335e-06,
+      "loss": 0.1972,
+      "step": 7480
+    },
+    {
+      "epoch": 0.09869979470442701,
+      "grad_norm": 1.2930989265441895,
+      "learning_rate": 9.876757847360516e-06,
+      "loss": 0.2136,
+      "step": 7500
+    },
+    {
+      "epoch": 0.09869979470442701,
+      "eval_loss": 0.2417767345905304,
+      "eval_runtime": 4.1765,
+      "eval_samples_per_second": 3.831,
+      "eval_steps_per_second": 1.915,
+      "step": 7500
+    },
+    {
+      "epoch": 0.09896299415697216,
+      "grad_norm": 1.0919674634933472,
+      "learning_rate": 9.875815590153356e-06,
+      "loss": 0.223,
+      "step": 7520
+    },
+    {
+      "epoch": 0.09922619360951729,
+      "grad_norm": 1.8348195552825928,
+      "learning_rate": 9.874869789892553e-06,
+      "loss": 0.2105,
+      "step": 7540
+    },
+    {
+      "epoch": 0.09948939306206243,
+      "grad_norm": 1.3600455522537231,
+      "learning_rate": 9.873920447265379e-06,
+      "loss": 0.2041,
+      "step": 7560
+    },
+    {
+      "epoch": 0.09975259251460757,
+      "grad_norm": 1.5351343154907227,
+      "learning_rate": 9.872967562961687e-06,
+      "loss": 0.2276,
+      "step": 7580
+    },
+    {
+      "epoch": 0.10001579196715271,
+      "grad_norm": 1.4131611585617065,
+      "learning_rate": 9.872011137673894e-06,
+      "loss": 0.2011,
+      "step": 7600
+    },
+    {
+      "epoch": 0.10027899141969784,
+      "grad_norm": 2.164309024810791,
+      "learning_rate": 9.871051172096997e-06,
+      "loss": 0.2338,
+      "step": 7620
+    },
+    {
+      "epoch": 0.10054219087224299,
+      "grad_norm": 1.3561838865280151,
+      "learning_rate": 9.870087666928562e-06,
+      "loss": 0.1995,
+      "step": 7640
+    },
+    {
+      "epoch": 0.10080539032478812,
+      "grad_norm": 1.5304696559906006,
+      "learning_rate": 9.86912062286873e-06,
+      "loss": 0.2197,
+      "step": 7660
+    },
+    {
+      "epoch": 0.10106858977733327,
+      "grad_norm": 1.3457480669021606,
+      "learning_rate": 9.86815004062021e-06,
+      "loss": 0.1971,
+      "step": 7680
+    },
+    {
+      "epoch": 0.1013317892298784,
+      "grad_norm": 2.637659788131714,
+      "learning_rate": 9.867175920888285e-06,
+      "loss": 0.215,
+      "step": 7700
+    },
+    {
+      "epoch": 0.10159498868242355,
+      "grad_norm": 1.8258278369903564,
+      "learning_rate": 9.86619826438081e-06,
+      "loss": 0.2122,
+      "step": 7720
+    },
+    {
+      "epoch": 0.10185818813496868,
+      "grad_norm": 1.4848147630691528,
+      "learning_rate": 9.865217071808202e-06,
+      "loss": 0.1999,
+      "step": 7740
+    },
+    {
+      "epoch": 0.10212138758751382,
+      "grad_norm": 1.7314472198486328,
+      "learning_rate": 9.86423234388346e-06,
+      "loss": 0.2245,
+      "step": 7760
+    },
+    {
+      "epoch": 0.10238458704005896,
+      "grad_norm": 1.5150028467178345,
+      "learning_rate": 9.863244081322137e-06,
+      "loss": 0.1993,
+      "step": 7780
+    },
+    {
+      "epoch": 0.1026477864926041,
+      "grad_norm": 1.582423448562622,
+      "learning_rate": 9.86225228484237e-06,
+      "loss": 0.2117,
+      "step": 7800
+    },
+    {
+      "epoch": 0.10291098594514923,
+      "grad_norm": 2.4957313537597656,
+      "learning_rate": 9.861256955164851e-06,
+      "loss": 0.2279,
+      "step": 7820
+    },
+    {
+      "epoch": 0.10317418539769438,
+      "grad_norm": 1.5010634660720825,
+      "learning_rate": 9.860258093012848e-06,
+      "loss": 0.2174,
+      "step": 7840
+    },
+    {
+      "epoch": 0.10343738485023951,
+      "grad_norm": 1.1131404638290405,
+      "learning_rate": 9.859255699112193e-06,
+      "loss": 0.205,
+      "step": 7860
+    },
+    {
+      "epoch": 0.10370058430278466,
+      "grad_norm": 1.3423928022384644,
+      "learning_rate": 9.858249774191284e-06,
+      "loss": 0.2065,
+      "step": 7880
+    },
+    {
+      "epoch": 0.10396378375532979,
+      "grad_norm": 1.2594897747039795,
+      "learning_rate": 9.857240318981084e-06,
+      "loss": 0.2167,
+      "step": 7900
+    },
+    {
+      "epoch": 0.10422698320787492,
+      "grad_norm": 1.3468621969223022,
+      "learning_rate": 9.856227334215123e-06,
+      "loss": 0.2184,
+      "step": 7920
+    },
+    {
+      "epoch": 0.10449018266042007,
+      "grad_norm": 1.4651366472244263,
+      "learning_rate": 9.855210820629495e-06,
+      "loss": 0.2128,
+      "step": 7940
+    },
+    {
+      "epoch": 0.1047533821129652,
+      "grad_norm": 1.729823112487793,
+      "learning_rate": 9.854190778962856e-06,
+      "loss": 0.1989,
+      "step": 7960
+    },
+    {
+      "epoch": 0.10501658156551034,
+      "grad_norm": 1.5334323644638062,
+      "learning_rate": 9.853167209956432e-06,
+      "loss": 0.2205,
+      "step": 7980
+    },
+    {
+      "epoch": 0.10527978101805548,
+      "grad_norm": 1.775930404663086,
+      "learning_rate": 9.852140114354008e-06,
+      "loss": 0.2064,
+      "step": 8000
+    },
+    {
+      "epoch": 0.10527978101805548,
+      "eval_loss": 0.2376280426979065,
+      "eval_runtime": 4.2711,
+      "eval_samples_per_second": 3.746,
+      "eval_steps_per_second": 1.873,
+      "step": 8000
+    },
+    {
+      "epoch": 0.10554298047060062,
+      "grad_norm": 1.3899028301239014,
+      "learning_rate": 9.851109492901929e-06,
+      "loss": 0.1956,
+      "step": 8020
+    },
+    {
+      "epoch": 0.10580617992314575,
+      "grad_norm": 2.1266427040100098,
+      "learning_rate": 9.850075346349108e-06,
+      "loss": 0.2053,
+      "step": 8040
+    },
+    {
+      "epoch": 0.1060693793756909,
+      "grad_norm": 1.9367544651031494,
+      "learning_rate": 9.849037675447014e-06,
+      "loss": 0.2081,
+      "step": 8060
+    },
+    {
+      "epoch": 0.10633257882823603,
+      "grad_norm": 2.0202436447143555,
+      "learning_rate": 9.847996480949682e-06,
+      "loss": 0.2225,
+      "step": 8080
+    },
+    {
+      "epoch": 0.10659577828078118,
+      "grad_norm": 1.5632342100143433,
+      "learning_rate": 9.846951763613704e-06,
+      "loss": 0.2039,
+      "step": 8100
+    },
+    {
+      "epoch": 0.10685897773332631,
+      "grad_norm": 1.5462706089019775,
+      "learning_rate": 9.84590352419823e-06,
+      "loss": 0.2199,
+      "step": 8120
+    },
+    {
+      "epoch": 0.10712217718587146,
+      "grad_norm": 1.1537566184997559,
+      "learning_rate": 9.844851763464977e-06,
+      "loss": 0.2117,
+      "step": 8140
+    },
+    {
+      "epoch": 0.10738537663841659,
+      "grad_norm": 1.1379863023757935,
+      "learning_rate": 9.843796482178214e-06,
+      "loss": 0.2338,
+      "step": 8160
+    },
+    {
+      "epoch": 0.10764857609096173,
+      "grad_norm": 1.560293197631836,
+      "learning_rate": 9.84273768110477e-06,
+      "loss": 0.2209,
+      "step": 8180
+    },
+    {
+      "epoch": 0.10791177554350687,
+      "grad_norm": 1.2221722602844238,
+      "learning_rate": 9.841675361014034e-06,
+      "loss": 0.2126,
+      "step": 8200
+    },
+    {
+      "epoch": 0.10817497499605201,
+      "grad_norm": 1.4082642793655396,
+      "learning_rate": 9.840609522677946e-06,
+      "loss": 0.2118,
+      "step": 8220
+    },
+    {
+      "epoch": 0.10843817444859714,
+      "grad_norm": 1.2365800142288208,
+      "learning_rate": 9.839540166871009e-06,
+      "loss": 0.2276,
+      "step": 8240
+    },
+    {
+      "epoch": 0.10870137390114229,
+      "grad_norm": 1.1324734687805176,
+      "learning_rate": 9.838467294370282e-06,
+      "loss": 0.1983,
+      "step": 8260
+    },
+    {
+      "epoch": 0.10896457335368742,
+      "grad_norm": 1.447879672050476,
+      "learning_rate": 9.837390905955372e-06,
+      "loss": 0.2257,
+      "step": 8280
+    },
+    {
+      "epoch": 0.10922777280623257,
+      "grad_norm": 1.8641455173492432,
+      "learning_rate": 9.836311002408451e-06,
+      "loss": 0.2194,
+      "step": 8300
+    },
+    {
+      "epoch": 0.1094909722587777,
+      "grad_norm": 1.5579819679260254,
+      "learning_rate": 9.835227584514236e-06,
+      "loss": 0.2198,
+      "step": 8320
+    },
+    {
+      "epoch": 0.10975417171132285,
+      "grad_norm": 1.513217806816101,
+      "learning_rate": 9.834140653060006e-06,
+      "loss": 0.2177,
+      "step": 8340
+    },
+    {
+      "epoch": 0.11001737116386798,
+      "grad_norm": 1.6074230670928955,
+      "learning_rate": 9.833050208835586e-06,
+      "loss": 0.2187,
+      "step": 8360
+    },
+    {
+      "epoch": 0.11028057061641312,
+      "grad_norm": 1.980845332145691,
+      "learning_rate": 9.83195625263336e-06,
+      "loss": 0.2101,
+      "step": 8380
+    },
+    {
+      "epoch": 0.11054377006895826,
+      "grad_norm": 1.2377039194107056,
+      "learning_rate": 9.830858785248254e-06,
+      "loss": 0.2127,
+      "step": 8400
+    },
+    {
+      "epoch": 0.1108069695215034,
+      "grad_norm": 1.3349093198776245,
+      "learning_rate": 9.829757807477762e-06,
+      "loss": 0.2129,
+      "step": 8420
+    },
+    {
+      "epoch": 0.11107016897404853,
+      "grad_norm": 1.2759006023406982,
+      "learning_rate": 9.82865332012191e-06,
+      "loss": 0.215,
+      "step": 8440
+    },
+    {
+      "epoch": 0.11133336842659368,
+      "grad_norm": 1.697147250175476,
+      "learning_rate": 9.827545323983289e-06,
+      "loss": 0.2127,
+      "step": 8460
+    },
+    {
+      "epoch": 0.11159656787913881,
+      "grad_norm": 1.4184967279434204,
+      "learning_rate": 9.826433819867032e-06,
+      "loss": 0.2143,
+      "step": 8480
+    },
+    {
+      "epoch": 0.11185976733168396,
+      "grad_norm": 1.3479217290878296,
+      "learning_rate": 9.825318808580824e-06,
+      "loss": 0.2094,
+      "step": 8500
+    },
+    {
+      "epoch": 0.11185976733168396,
+      "eval_loss": 0.24468787014484406,
+      "eval_runtime": 4.3364,
+      "eval_samples_per_second": 3.69,
+      "eval_steps_per_second": 1.845,
+      "step": 8500
+    },
+    {
+      "epoch": 0.11212296678422909,
+      "grad_norm": 1.3764429092407227,
+      "learning_rate": 9.824200290934896e-06,
+      "loss": 0.2072,
+      "step": 8520
+    },
+    {
+      "epoch": 0.11238616623677422,
+      "grad_norm": 1.8549425601959229,
+      "learning_rate": 9.82307826774203e-06,
+      "loss": 0.2202,
+      "step": 8540
+    },
+    {
+      "epoch": 0.11264936568931937,
+      "grad_norm": 1.870877981185913,
+      "learning_rate": 9.821952739817554e-06,
+      "loss": 0.2233,
+      "step": 8560
+    },
+    {
+      "epoch": 0.1129125651418645,
+      "grad_norm": 1.4307260513305664,
+      "learning_rate": 9.82082370797934e-06,
+      "loss": 0.2148,
+      "step": 8580
+    },
+    {
+      "epoch": 0.11317576459440964,
+      "grad_norm": 1.0387654304504395,
+      "learning_rate": 9.819691173047812e-06,
+      "loss": 0.2078,
+      "step": 8600
+    },
+    {
+      "epoch": 0.11343896404695478,
+      "grad_norm": 1.268905520439148,
+      "learning_rate": 9.818555135845934e-06,
+      "loss": 0.2196,
+      "step": 8620
+    },
+    {
+      "epoch": 0.11370216349949992,
+      "grad_norm": 1.4618275165557861,
+      "learning_rate": 9.817415597199219e-06,
+      "loss": 0.1805,
+      "step": 8640
+    },
+    {
+      "epoch": 0.11396536295204505,
+      "grad_norm": 1.4536092281341553,
+      "learning_rate": 9.816272557935721e-06,
+      "loss": 0.2071,
+      "step": 8660
+    },
+    {
+      "epoch": 0.1142285624045902,
+      "grad_norm": 1.5338225364685059,
+      "learning_rate": 9.815126018886043e-06,
+      "loss": 0.2034,
+      "step": 8680
+    },
+    {
+      "epoch": 0.11449176185713533,
+      "grad_norm": 1.230372428894043,
+      "learning_rate": 9.813975980883322e-06,
+      "loss": 0.1989,
+      "step": 8700
+    },
+    {
+      "epoch": 0.11475496130968048,
+      "grad_norm": 1.380357265472412,
+      "learning_rate": 9.812822444763249e-06,
+      "loss": 0.2214,
+      "step": 8720
+    },
+    {
+      "epoch": 0.11501816076222561,
+      "grad_norm": 1.3375415802001953,
+      "learning_rate": 9.811665411364043e-06,
+      "loss": 0.2141,
+      "step": 8740
+    },
+    {
+      "epoch": 0.11528136021477076,
+      "grad_norm": 1.4750745296478271,
+      "learning_rate": 9.810504881526482e-06,
+      "loss": 0.1957,
+      "step": 8760
+    },
+    {
+      "epoch": 0.11554455966731589,
+      "grad_norm": 1.6168025732040405,
+      "learning_rate": 9.80934085609387e-06,
+      "loss": 0.2098,
+      "step": 8780
+    },
+    {
+      "epoch": 0.11580775911986103,
+      "grad_norm": 1.6745901107788086,
+      "learning_rate": 9.808173335912054e-06,
+      "loss": 0.2211,
+      "step": 8800
+    },
+    {
+      "epoch": 0.11607095857240617,
+      "grad_norm": 1.3100727796554565,
+      "learning_rate": 9.807002321829426e-06,
+      "loss": 0.2094,
+      "step": 8820
+    },
+    {
+      "epoch": 0.11633415802495131,
+      "grad_norm": 1.2737947702407837,
+      "learning_rate": 9.805827814696913e-06,
+      "loss": 0.2062,
+      "step": 8840
+    },
+    {
+      "epoch": 0.11659735747749644,
+      "grad_norm": 1.3218148946762085,
+      "learning_rate": 9.80464981536798e-06,
+      "loss": 0.207,
+      "step": 8860
+    },
+    {
+      "epoch": 0.11686055693004159,
+      "grad_norm": 1.4089081287384033,
+      "learning_rate": 9.803468324698631e-06,
+      "loss": 0.2144,
+      "step": 8880
+    },
+    {
+      "epoch": 0.11712375638258672,
+      "grad_norm": 1.3648630380630493,
+      "learning_rate": 9.802283343547408e-06,
+      "loss": 0.1992,
+      "step": 8900
+    },
+    {
+      "epoch": 0.11738695583513187,
+      "grad_norm": 1.0359928607940674,
+      "learning_rate": 9.801094872775385e-06,
+      "loss": 0.2148,
+      "step": 8920
+    },
+    {
+      "epoch": 0.117650155287677,
+      "grad_norm": 1.1766061782836914,
+      "learning_rate": 9.799902913246175e-06,
+      "loss": 0.2055,
+      "step": 8940
+    },
+    {
+      "epoch": 0.11791335474022215,
+      "grad_norm": 1.4708646535873413,
+      "learning_rate": 9.798707465825928e-06,
+      "loss": 0.2096,
+      "step": 8960
+    },
+    {
+      "epoch": 0.11817655419276728,
+      "grad_norm": 1.3902626037597656,
+      "learning_rate": 9.797508531383323e-06,
+      "loss": 0.2053,
+      "step": 8980
+    },
+    {
+      "epoch": 0.11843975364531242,
+      "grad_norm": 1.569837212562561,
+      "learning_rate": 9.796306110789578e-06,
+      "loss": 0.2115,
+      "step": 9000
+    },
+    {
+      "epoch": 0.11843975364531242,
+      "eval_loss": 0.24664020538330078,
+      "eval_runtime": 4.3431,
+      "eval_samples_per_second": 3.684,
+      "eval_steps_per_second": 1.842,
+      "step": 9000
+    },
+    {
+      "epoch": 0.11870295309785756,
+      "grad_norm": 1.5649627447128296,
+      "learning_rate": 9.795100204918443e-06,
+      "loss": 0.2175,
+      "step": 9020
+    },
+    {
+      "epoch": 0.1189661525504027,
+      "grad_norm": 1.0380131006240845,
+      "learning_rate": 9.7938908146462e-06,
+      "loss": 0.233,
+      "step": 9040
+    },
+    {
+      "epoch": 0.11922935200294783,
+      "grad_norm": 0.9991648197174072,
+      "learning_rate": 9.79267794085166e-06,
+      "loss": 0.2124,
+      "step": 9060
+    },
+    {
+      "epoch": 0.11949255145549298,
+      "grad_norm": 1.2039517164230347,
+      "learning_rate": 9.791461584416172e-06,
+      "loss": 0.2237,
+      "step": 9080
+    },
+    {
+      "epoch": 0.11975575090803811,
+      "grad_norm": 1.0578422546386719,
+      "learning_rate": 9.790241746223608e-06,
+      "loss": 0.219,
+      "step": 9100
+    },
+    {
+      "epoch": 0.12001895036058324,
+      "grad_norm": 1.3310569524765015,
+      "learning_rate": 9.789018427160377e-06,
+      "loss": 0.2173,
+      "step": 9120
+    },
+    {
+      "epoch": 0.12028214981312839,
+      "grad_norm": 1.322142243385315,
+      "learning_rate": 9.787791628115414e-06,
+      "loss": 0.2236,
+      "step": 9140
+    },
+    {
+      "epoch": 0.12054534926567352,
+      "grad_norm": 1.667343020439148,
+      "learning_rate": 9.786561349980182e-06,
+      "loss": 0.1988,
+      "step": 9160
+    },
+    {
+      "epoch": 0.12080854871821867,
+      "grad_norm": 1.1755404472351074,
+      "learning_rate": 9.785327593648675e-06,
+      "loss": 0.2027,
+      "step": 9180
+    },
+    {
+      "epoch": 0.1210717481707638,
+      "grad_norm": 1.4779407978057861,
+      "learning_rate": 9.784090360017411e-06,
+      "loss": 0.2113,
+      "step": 9200
+    },
+    {
+      "epoch": 0.12133494762330894,
+      "grad_norm": 1.229215383529663,
+      "learning_rate": 9.782849649985438e-06,
+      "loss": 0.2259,
+      "step": 9220
+    },
+    {
+      "epoch": 0.12159814707585408,
+      "grad_norm": 1.4789106845855713,
+      "learning_rate": 9.781605464454328e-06,
+      "loss": 0.1983,
+      "step": 9240
+    },
+    {
+      "epoch": 0.12186134652839922,
+      "grad_norm": 1.0277549028396606,
+      "learning_rate": 9.780357804328178e-06,
+      "loss": 0.2107,
+      "step": 9260
+    },
+    {
+      "epoch": 0.12212454598094435,
+      "grad_norm": 1.3237489461898804,
+      "learning_rate": 9.779106670513614e-06,
+      "loss": 0.2303,
+      "step": 9280
+    },
+    {
+      "epoch": 0.1223877454334895,
+      "grad_norm": 1.7285518646240234,
+      "learning_rate": 9.77785206391978e-06,
+      "loss": 0.1977,
+      "step": 9300
+    },
+    {
+      "epoch": 0.12265094488603463,
+      "grad_norm": 1.4146068096160889,
+      "learning_rate": 9.77659398545835e-06,
+      "loss": 0.2195,
+      "step": 9320
+    },
+    {
+      "epoch": 0.12291414433857978,
+      "grad_norm": 1.1390304565429688,
+      "learning_rate": 9.775332436043514e-06,
+      "loss": 0.2074,
+      "step": 9340
+    },
+    {
+      "epoch": 0.12317734379112491,
+      "grad_norm": 1.4106401205062866,
+      "learning_rate": 9.774067416591988e-06,
+      "loss": 0.222,
+      "step": 9360
+    },
+    {
+      "epoch": 0.12344054324367006,
+      "grad_norm": 1.1507915258407593,
+      "learning_rate": 9.772798928023012e-06,
+      "loss": 0.2099,
+      "step": 9380
+    },
+    {
+      "epoch": 0.12370374269621519,
+      "grad_norm": 1.7786073684692383,
+      "learning_rate": 9.771526971258344e-06,
+      "loss": 0.2273,
+      "step": 9400
+    },
+    {
+      "epoch": 0.12396694214876033,
+      "grad_norm": 1.7390908002853394,
+      "learning_rate": 9.770251547222262e-06,
+      "loss": 0.2057,
+      "step": 9420
+    },
+    {
+      "epoch": 0.12423014160130547,
+      "grad_norm": 1.1683825254440308,
+      "learning_rate": 9.768972656841564e-06,
+      "loss": 0.2265,
+      "step": 9440
+    },
+    {
+      "epoch": 0.12449334105385061,
+      "grad_norm": 1.6143608093261719,
+      "learning_rate": 9.767690301045564e-06,
+      "loss": 0.2062,
+      "step": 9460
+    },
+    {
+      "epoch": 0.12475654050639574,
+      "grad_norm": 2.3575456142425537,
+      "learning_rate": 9.766404480766102e-06,
+      "loss": 0.1953,
+      "step": 9480
+    },
+    {
+      "epoch": 0.1250197399589409,
+      "grad_norm": 1.2589553594589233,
+      "learning_rate": 9.765115196937527e-06,
+      "loss": 0.2177,
+      "step": 9500
+    },
+    {
+      "epoch": 0.1250197399589409,
+      "eval_loss": 0.23723182082176208,
+      "eval_runtime": 3.5043,
+      "eval_samples_per_second": 4.566,
+      "eval_steps_per_second": 2.283,
+      "step": 9500
+    },
+    {
+      "epoch": 0.12528293941148602,
+      "grad_norm": 1.6730248928070068,
+      "learning_rate": 9.763822450496707e-06,
+      "loss": 0.1985,
+      "step": 9520
+    },
+    {
+      "epoch": 0.12554613886403115,
+      "grad_norm": 1.455782413482666,
+      "learning_rate": 9.762526242383031e-06,
+      "loss": 0.2106,
+      "step": 9540
+    },
+    {
+      "epoch": 0.1258093383165763,
+      "grad_norm": 1.4631160497665405,
+      "learning_rate": 9.761226573538397e-06,
+      "loss": 0.2145,
+      "step": 9560
+    },
+    {
+      "epoch": 0.12607253776912145,
+      "grad_norm": 1.6533704996109009,
+      "learning_rate": 9.759923444907223e-06,
+      "loss": 0.2162,
+      "step": 9580
+    },
+    {
+      "epoch": 0.12633573722166658,
+      "grad_norm": 1.3942186832427979,
+      "learning_rate": 9.758616857436435e-06,
+      "loss": 0.2201,
+      "step": 9600
+    },
+    {
+      "epoch": 0.1265989366742117,
+      "grad_norm": 1.3460062742233276,
+      "learning_rate": 9.757306812075483e-06,
+      "loss": 0.2278,
+      "step": 9620
+    },
+    {
+      "epoch": 0.12686213612675687,
+      "grad_norm": 1.401207685470581,
+      "learning_rate": 9.755993309776315e-06,
+      "loss": 0.2088,
+      "step": 9640
+    },
+    {
+      "epoch": 0.127125335579302,
+      "grad_norm": 1.607262372970581,
+      "learning_rate": 9.754676351493402e-06,
+      "loss": 0.1997,
+      "step": 9660
+    },
+    {
+      "epoch": 0.12738853503184713,
+      "grad_norm": 1.3810948133468628,
+      "learning_rate": 9.753355938183725e-06,
+      "loss": 0.2168,
+      "step": 9680
+    },
+    {
+      "epoch": 0.12765173448439227,
+      "grad_norm": 1.351562261581421,
+      "learning_rate": 9.752032070806772e-06,
+      "loss": 0.2235,
+      "step": 9700
+    },
+    {
+      "epoch": 0.12791493393693742,
+      "grad_norm": 1.519513487815857,
+      "learning_rate": 9.750704750324542e-06,
+      "loss": 0.1878,
+      "step": 9720
+    },
+    {
+      "epoch": 0.12817813338948256,
+      "grad_norm": 1.138999581336975,
+      "learning_rate": 9.749373977701547e-06,
+      "loss": 0.2252,
+      "step": 9740
+    },
+    {
+      "epoch": 0.1284413328420277,
+      "grad_norm": 1.3984180688858032,
+      "learning_rate": 9.7480397539048e-06,
+      "loss": 0.1942,
+      "step": 9760
+    },
+    {
+      "epoch": 0.12870453229457282,
+      "grad_norm": 1.1762109994888306,
+      "learning_rate": 9.74670207990383e-06,
+      "loss": 0.2192,
+      "step": 9780
+    },
+    {
+      "epoch": 0.12896773174711795,
+      "grad_norm": 1.8021773099899292,
+      "learning_rate": 9.745360956670666e-06,
+      "loss": 0.2096,
+      "step": 9800
+    },
+    {
+      "epoch": 0.1292309311996631,
+      "grad_norm": 1.4322994947433472,
+      "learning_rate": 9.744016385179851e-06,
+      "loss": 0.205,
+      "step": 9820
+    },
+    {
+      "epoch": 0.12949413065220824,
+      "grad_norm": 2.472860336303711,
+      "learning_rate": 9.742668366408425e-06,
+      "loss": 0.2003,
+      "step": 9840
+    },
+    {
+      "epoch": 0.12975733010475338,
+      "grad_norm": 1.4821882247924805,
+      "learning_rate": 9.741316901335941e-06,
+      "loss": 0.214,
+      "step": 9860
+    },
+    {
+      "epoch": 0.1300205295572985,
+      "grad_norm": 1.400039792060852,
+      "learning_rate": 9.739961990944451e-06,
+      "loss": 0.2088,
+      "step": 9880
+    },
+    {
+      "epoch": 0.13028372900984367,
+      "grad_norm": 1.5771602392196655,
+      "learning_rate": 9.738603636218514e-06,
+      "loss": 0.1944,
+      "step": 9900
+    },
+    {
+      "epoch": 0.1305469284623888,
+      "grad_norm": 1.8049954175949097,
+      "learning_rate": 9.737241838145189e-06,
+      "loss": 0.2206,
+      "step": 9920
+    },
+    {
+      "epoch": 0.13081012791493393,
+      "grad_norm": 1.9996815919876099,
+      "learning_rate": 9.735876597714038e-06,
+      "loss": 0.2212,
+      "step": 9940
+    },
+    {
+      "epoch": 0.13107332736747906,
+      "grad_norm": 1.5141178369522095,
+      "learning_rate": 9.734507915917126e-06,
+      "loss": 0.2058,
+      "step": 9960
+    },
+    {
+      "epoch": 0.13133652682002422,
+      "grad_norm": 1.7168971300125122,
+      "learning_rate": 9.733135793749017e-06,
+      "loss": 0.1931,
+      "step": 9980
+    },
+    {
+      "epoch": 0.13159972627256936,
+      "grad_norm": 1.4730061292648315,
+      "learning_rate": 9.731760232206775e-06,
+      "loss": 0.1995,
+      "step": 10000
+    },
+    {
+      "epoch": 0.13159972627256936,
+      "eval_loss": 0.2411240041255951,
+      "eval_runtime": 4.2261,
+      "eval_samples_per_second": 3.786,
+      "eval_steps_per_second": 1.893,
+      "step": 10000
+    },
+    {
+      "epoch": 0.1318629257251145,
+      "grad_norm": 1.4055002927780151,
+      "learning_rate": 9.730381232289964e-06,
+      "loss": 0.221,
+      "step": 10020
+    },
+    {
+      "epoch": 0.13212612517765962,
+      "grad_norm": 1.595478892326355,
+      "learning_rate": 9.728998795000646e-06,
+      "loss": 0.2266,
+      "step": 10040
+    },
+    {
+      "epoch": 0.13238932463020478,
+      "grad_norm": 1.2498458623886108,
+      "learning_rate": 9.72761292134338e-06,
+      "loss": 0.2101,
+      "step": 10060
+    },
+    {
+      "epoch": 0.1326525240827499,
+      "grad_norm": 1.2517942190170288,
+      "learning_rate": 9.726223612325227e-06,
+      "loss": 0.2143,
+      "step": 10080
+    },
+    {
+      "epoch": 0.13291572353529504,
+      "grad_norm": 2.1070661544799805,
+      "learning_rate": 9.724830868955736e-06,
+      "loss": 0.2045,
+      "step": 10100
+    },
+    {
+      "epoch": 0.13317892298784018,
+      "grad_norm": 1.237102746963501,
+      "learning_rate": 9.723434692246958e-06,
+      "loss": 0.2005,
+      "step": 10120
+    },
+    {
+      "epoch": 0.13344212244038534,
+      "grad_norm": 1.5408658981323242,
+      "learning_rate": 9.722035083213438e-06,
+      "loss": 0.2107,
+      "step": 10140
+    },
+    {
+      "epoch": 0.13370532189293047,
+      "grad_norm": 1.307072401046753,
+      "learning_rate": 9.720632042872214e-06,
+      "loss": 0.1965,
+      "step": 10160
+    },
+    {
+      "epoch": 0.1339685213454756,
+      "grad_norm": 1.4514399766921997,
+      "learning_rate": 9.719225572242813e-06,
+      "loss": 0.2018,
+      "step": 10180
+    },
+    {
+      "epoch": 0.13423172079802073,
+      "grad_norm": 1.5330537557601929,
+      "learning_rate": 9.717815672347264e-06,
+      "loss": 0.2135,
+      "step": 10200
+    },
+    {
+      "epoch": 0.1344949202505659,
+      "grad_norm": 1.133673071861267,
+      "learning_rate": 9.716402344210081e-06,
+      "loss": 0.2177,
+      "step": 10220
+    },
+    {
+      "epoch": 0.13475811970311102,
+      "grad_norm": 1.4490848779678345,
+      "learning_rate": 9.71498558885827e-06,
+      "loss": 0.2108,
+      "step": 10240
+    },
+    {
+      "epoch": 0.13502131915565616,
+      "grad_norm": 1.5667150020599365,
+      "learning_rate": 9.713565407321333e-06,
+      "loss": 0.2002,
+      "step": 10260
+    },
+    {
+      "epoch": 0.1352845186082013,
+      "grad_norm": 1.62337064743042,
+      "learning_rate": 9.712141800631251e-06,
+      "loss": 0.212,
+      "step": 10280
+    },
+    {
+      "epoch": 0.13554771806074645,
+      "grad_norm": 1.416542410850525,
+      "learning_rate": 9.710714769822506e-06,
+      "loss": 0.2109,
+      "step": 10300
+    },
+    {
+      "epoch": 0.13581091751329158,
+      "grad_norm": 1.3092331886291504,
+      "learning_rate": 9.709284315932058e-06,
+      "loss": 0.192,
+      "step": 10320
+    },
+    {
+      "epoch": 0.1360741169658367,
+      "grad_norm": 1.4110162258148193,
+      "learning_rate": 9.70785043999936e-06,
+      "loss": 0.2134,
+      "step": 10340
+    },
+    {
+      "epoch": 0.13633731641838184,
+      "grad_norm": 2.1192452907562256,
+      "learning_rate": 9.706413143066351e-06,
+      "loss": 0.2084,
+      "step": 10360
+    },
+    {
+      "epoch": 0.13660051587092698,
+      "grad_norm": 1.2066258192062378,
+      "learning_rate": 9.704972426177458e-06,
+      "loss": 0.1973,
+      "step": 10380
+    },
+    {
+      "epoch": 0.13686371532347213,
+      "grad_norm": 1.8961396217346191,
+      "learning_rate": 9.703528290379582e-06,
+      "loss": 0.2129,
+      "step": 10400
+    },
+    {
+      "epoch": 0.13712691477601727,
+      "grad_norm": 1.8682118654251099,
+      "learning_rate": 9.702080736722125e-06,
+      "loss": 0.2192,
+      "step": 10420
+    },
+    {
+      "epoch": 0.1373901142285624,
+      "grad_norm": 1.46285080909729,
+      "learning_rate": 9.70062976625696e-06,
+      "loss": 0.2129,
+      "step": 10440
+    },
+    {
+      "epoch": 0.13765331368110753,
+      "grad_norm": 1.1597774028778076,
+      "learning_rate": 9.699175380038452e-06,
+      "loss": 0.217,
+      "step": 10460
+    },
+    {
+      "epoch": 0.1379165131336527,
+      "grad_norm": 1.7788597345352173,
+      "learning_rate": 9.69771757912344e-06,
+      "loss": 0.2123,
+      "step": 10480
+    },
+    {
+      "epoch": 0.13817971258619782,
+      "grad_norm": 1.252797245979309,
+      "learning_rate": 9.696256364571246e-06,
+      "loss": 0.2157,
+      "step": 10500
+    },
+    {
+      "epoch": 0.13817971258619782,
+      "eval_loss": 0.2348114252090454,
+      "eval_runtime": 3.9521,
+      "eval_samples_per_second": 4.048,
+      "eval_steps_per_second": 2.024,
+      "step": 10500
+    },
+    {
+      "epoch": 0.13844291203874295,
+      "grad_norm": 1.74348783493042,
+      "learning_rate": 9.694791737443677e-06,
+      "loss": 0.2342,
+      "step": 10520
+    },
+    {
+      "epoch": 0.1387061114912881,
+      "grad_norm": 1.5580233335494995,
+      "learning_rate": 9.693323698805014e-06,
+      "loss": 0.2058,
+      "step": 10540
+    },
+    {
+      "epoch": 0.13896931094383325,
+      "grad_norm": 1.4080278873443604,
+      "learning_rate": 9.691852249722023e-06,
+      "loss": 0.2127,
+      "step": 10560
+    },
+    {
+      "epoch": 0.13923251039637838,
+      "grad_norm": 1.5714479684829712,
+      "learning_rate": 9.690377391263946e-06,
+      "loss": 0.199,
+      "step": 10580
+    },
+    {
+      "epoch": 0.1394957098489235,
+      "grad_norm": 1.6956638097763062,
+      "learning_rate": 9.688899124502498e-06,
+      "loss": 0.2116,
+      "step": 10600
+    },
+    {
+      "epoch": 0.13975890930146864,
+      "grad_norm": 1.4666132926940918,
+      "learning_rate": 9.687417450511876e-06,
+      "loss": 0.1912,
+      "step": 10620
+    },
+    {
+      "epoch": 0.1400221087540138,
+      "grad_norm": 1.2092963457107544,
+      "learning_rate": 9.685932370368752e-06,
+      "loss": 0.2036,
+      "step": 10640
+    },
+    {
+      "epoch": 0.14028530820655893,
+      "grad_norm": 1.233964204788208,
+      "learning_rate": 9.684443885152272e-06,
+      "loss": 0.2106,
+      "step": 10660
+    },
+    {
+      "epoch": 0.14054850765910407,
+      "grad_norm": 1.4183176755905151,
+      "learning_rate": 9.682951995944058e-06,
+      "loss": 0.2152,
+      "step": 10680
+    },
+    {
+      "epoch": 0.1408117071116492,
+      "grad_norm": 1.4814904928207397,
+      "learning_rate": 9.681456703828202e-06,
+      "loss": 0.2333,
+      "step": 10700
+    },
+    {
+      "epoch": 0.14107490656419436,
+      "grad_norm": 1.6026979684829712,
+      "learning_rate": 9.67995800989127e-06,
+      "loss": 0.214,
+      "step": 10720
+    },
+    {
+      "epoch": 0.1413381060167395,
+      "grad_norm": 1.328474521636963,
+      "learning_rate": 9.678455915222304e-06,
+      "loss": 0.1964,
+      "step": 10740
+    },
+    {
+      "epoch": 0.14160130546928462,
+      "grad_norm": 1.1682385206222534,
+      "learning_rate": 9.676950420912814e-06,
+      "loss": 0.2252,
+      "step": 10760
+    },
+    {
+      "epoch": 0.14186450492182975,
+      "grad_norm": 1.6274493932724,
+      "learning_rate": 9.675441528056778e-06,
+      "loss": 0.1976,
+      "step": 10780
+    },
+    {
+      "epoch": 0.1421277043743749,
+      "grad_norm": 1.0832542181015015,
+      "learning_rate": 9.673929237750647e-06,
+      "loss": 0.2214,
+      "step": 10800
+    },
+    {
+      "epoch": 0.14239090382692005,
+      "grad_norm": 2.0469839572906494,
+      "learning_rate": 9.672413551093342e-06,
+      "loss": 0.2151,
+      "step": 10820
+    },
+    {
+      "epoch": 0.14265410327946518,
+      "grad_norm": 1.3626993894577026,
+      "learning_rate": 9.670894469186247e-06,
+      "loss": 0.2099,
+      "step": 10840
+    },
+    {
+      "epoch": 0.1429173027320103,
+      "grad_norm": 1.7062361240386963,
+      "learning_rate": 9.669371993133217e-06,
+      "loss": 0.2061,
+      "step": 10860
+    },
+    {
+      "epoch": 0.14318050218455547,
+      "grad_norm": 1.2700859308242798,
+      "learning_rate": 9.667846124040572e-06,
+      "loss": 0.2156,
+      "step": 10880
+    },
+    {
+      "epoch": 0.1434437016371006,
+      "grad_norm": 2.135843276977539,
+      "learning_rate": 9.666316863017097e-06,
+      "loss": 0.2019,
+      "step": 10900
+    },
+    {
+      "epoch": 0.14370690108964573,
+      "grad_norm": 1.8414427042007446,
+      "learning_rate": 9.664784211174047e-06,
+      "loss": 0.214,
+      "step": 10920
+    },
+    {
+      "epoch": 0.14397010054219087,
+      "grad_norm": 1.6474508047103882,
+      "learning_rate": 9.663248169625133e-06,
+      "loss": 0.201,
+      "step": 10940
+    },
+    {
+      "epoch": 0.144233299994736,
+      "grad_norm": 1.9574576616287231,
+      "learning_rate": 9.661708739486534e-06,
+      "loss": 0.195,
+      "step": 10960
+    },
+    {
+      "epoch": 0.14449649944728116,
+      "grad_norm": 1.7416417598724365,
+      "learning_rate": 9.660165921876889e-06,
+      "loss": 0.2055,
+      "step": 10980
+    },
+    {
+      "epoch": 0.1447596988998263,
+      "grad_norm": 1.4388797283172607,
+      "learning_rate": 9.658619717917304e-06,
+      "loss": 0.2044,
+      "step": 11000
+    },
+    {
+      "epoch": 0.1447596988998263,
+      "eval_loss": 0.24022623896598816,
+      "eval_runtime": 4.3018,
+      "eval_samples_per_second": 3.719,
+      "eval_steps_per_second": 1.86,
+      "step": 11000
+    },
+    {
+      "epoch": 0.14502289835237142,
+      "grad_norm": 2.4044790267944336,
+      "learning_rate": 9.657070128731336e-06,
+      "loss": 0.2238,
+      "step": 11020
+    },
+    {
+      "epoch": 0.14528609780491655,
+      "grad_norm": 1.6349101066589355,
+      "learning_rate": 9.655517155445009e-06,
+      "loss": 0.2161,
+      "step": 11040
+    },
+    {
+      "epoch": 0.1455492972574617,
+      "grad_norm": 1.867083191871643,
+      "learning_rate": 9.653960799186807e-06,
+      "loss": 0.2197,
+      "step": 11060
+    },
+    {
+      "epoch": 0.14581249671000684,
+      "grad_norm": 1.4095635414123535,
+      "learning_rate": 9.65240106108767e-06,
+      "loss": 0.2038,
+      "step": 11080
+    },
+    {
+      "epoch": 0.14607569616255198,
+      "grad_norm": 1.6474804878234863,
+      "learning_rate": 9.650837942280988e-06,
+      "loss": 0.2044,
+      "step": 11100
+    },
+    {
+      "epoch": 0.1463388956150971,
+      "grad_norm": 1.4282419681549072,
+      "learning_rate": 9.649271443902624e-06,
+      "loss": 0.1959,
+      "step": 11120
+    },
+    {
+      "epoch": 0.14660209506764227,
+      "grad_norm": 1.444464921951294,
+      "learning_rate": 9.647701567090884e-06,
+      "loss": 0.1976,
+      "step": 11140
+    },
+    {
+      "epoch": 0.1468652945201874,
+      "grad_norm": 1.3266487121582031,
+      "learning_rate": 9.64612831298653e-06,
+      "loss": 0.2136,
+      "step": 11160
+    },
+    {
+      "epoch": 0.14712849397273253,
+      "grad_norm": 1.1017826795578003,
+      "learning_rate": 9.644551682732784e-06,
+      "loss": 0.2367,
+      "step": 11180
+    },
+    {
+      "epoch": 0.14739169342527766,
+      "grad_norm": 1.8925037384033203,
+      "learning_rate": 9.642971677475316e-06,
+      "loss": 0.2185,
+      "step": 11200
+    },
+    {
+      "epoch": 0.14765489287782282,
+      "grad_norm": 1.4914782047271729,
+      "learning_rate": 9.641388298362254e-06,
+      "loss": 0.1996,
+      "step": 11220
+    },
+    {
+      "epoch": 0.14791809233036796,
+      "grad_norm": 1.4451285600662231,
+      "learning_rate": 9.63980154654417e-06,
+      "loss": 0.2035,
+      "step": 11240
+    },
+    {
+      "epoch": 0.1481812917829131,
+      "grad_norm": 1.314845323562622,
+      "learning_rate": 9.638211423174095e-06,
+      "loss": 0.217,
+      "step": 11260
+    },
+    {
+      "epoch": 0.14844449123545822,
+      "grad_norm": 1.5677908658981323,
+      "learning_rate": 9.636617929407501e-06,
+      "loss": 0.2143,
+      "step": 11280
+    },
+    {
+      "epoch": 0.14870769068800338,
+      "grad_norm": 1.344436764717102,
+      "learning_rate": 9.635021066402319e-06,
+      "loss": 0.1958,
+      "step": 11300
+    },
+    {
+      "epoch": 0.1489708901405485,
+      "grad_norm": 1.3840956687927246,
+      "learning_rate": 9.633420835318918e-06,
+      "loss": 0.2082,
+      "step": 11320
+    },
+    {
+      "epoch": 0.14923408959309364,
+      "grad_norm": 1.548115849494934,
+      "learning_rate": 9.631817237320124e-06,
+      "loss": 0.2139,
+      "step": 11340
+    },
+    {
+      "epoch": 0.14949728904563878,
+      "grad_norm": 1.9579691886901855,
+      "learning_rate": 9.630210273571205e-06,
+      "loss": 0.2017,
+      "step": 11360
+    },
+    {
+      "epoch": 0.14976048849818394,
+      "grad_norm": 1.8819093704223633,
+      "learning_rate": 9.628599945239874e-06,
+      "loss": 0.2081,
+      "step": 11380
+    },
+    {
+      "epoch": 0.15002368795072907,
+      "grad_norm": 1.5918498039245605,
+      "learning_rate": 9.626986253496291e-06,
+      "loss": 0.1989,
+      "step": 11400
+    },
+    {
+      "epoch": 0.1502868874032742,
+      "grad_norm": 1.9715248346328735,
+      "learning_rate": 9.625369199513057e-06,
+      "loss": 0.2127,
+      "step": 11420
+    },
+    {
+      "epoch": 0.15055008685581933,
+      "grad_norm": 1.7663583755493164,
+      "learning_rate": 9.62374878446522e-06,
+      "loss": 0.2231,
+      "step": 11440
+    },
+    {
+      "epoch": 0.1508132863083645,
+      "grad_norm": 1.4816879034042358,
+      "learning_rate": 9.62212500953027e-06,
+      "loss": 0.2042,
+      "step": 11460
+    },
+    {
+      "epoch": 0.15107648576090962,
+      "grad_norm": 1.305472731590271,
+      "learning_rate": 9.620497875888133e-06,
+      "loss": 0.2054,
+      "step": 11480
+    },
+    {
+      "epoch": 0.15133968521345476,
+      "grad_norm": 1.3565542697906494,
+      "learning_rate": 9.618867384721184e-06,
+      "loss": 0.2023,
+      "step": 11500
+    },
+    {
+      "epoch": 0.15133968521345476,
+      "eval_loss": 0.23877675831317902,
+      "eval_runtime": 4.2955,
+      "eval_samples_per_second": 3.725,
+      "eval_steps_per_second": 1.862,
+      "step": 11500
+    },
+    {
+      "epoch": 0.1516028846659999,
+      "grad_norm": 1.2534278631210327,
+      "learning_rate": 9.617233537214232e-06,
+      "loss": 0.2197,
+      "step": 11520
+    },
+    {
+      "epoch": 0.15186608411854505,
+      "grad_norm": 1.1261582374572754,
+      "learning_rate": 9.615596334554525e-06,
+      "loss": 0.2263,
+      "step": 11540
+    },
+    {
+      "epoch": 0.15212928357109018,
+      "grad_norm": 1.4619392156600952,
+      "learning_rate": 9.613955777931752e-06,
+      "loss": 0.1942,
+      "step": 11560
+    },
+    {
+      "epoch": 0.1523924830236353,
+      "grad_norm": 1.6779810190200806,
+      "learning_rate": 9.612311868538038e-06,
+      "loss": 0.1975,
+      "step": 11580
+    },
+    {
+      "epoch": 0.15265568247618044,
+      "grad_norm": 1.611639142036438,
+      "learning_rate": 9.610664607567942e-06,
+      "loss": 0.2056,
+      "step": 11600
+    },
+    {
+      "epoch": 0.15291888192872558,
+      "grad_norm": 1.4214487075805664,
+      "learning_rate": 9.609013996218463e-06,
+      "loss": 0.234,
+      "step": 11620
+    },
+    {
+      "epoch": 0.15318208138127073,
+      "grad_norm": 1.6175751686096191,
+      "learning_rate": 9.607360035689031e-06,
+      "loss": 0.2034,
+      "step": 11640
+    },
+    {
+      "epoch": 0.15344528083381587,
+      "grad_norm": 1.081208348274231,
+      "learning_rate": 9.60570272718151e-06,
+      "loss": 0.1942,
+      "step": 11660
+    },
+    {
+      "epoch": 0.153708480286361,
+      "grad_norm": 1.3809984922409058,
+      "learning_rate": 9.604042071900198e-06,
+      "loss": 0.2123,
+      "step": 11680
+    },
+    {
+      "epoch": 0.15397167973890613,
+      "grad_norm": 1.1256625652313232,
+      "learning_rate": 9.602378071051825e-06,
+      "loss": 0.2398,
+      "step": 11700
+    },
+    {
+      "epoch": 0.1542348791914513,
+      "grad_norm": 1.3902660608291626,
+      "learning_rate": 9.600710725845552e-06,
+      "loss": 0.2165,
+      "step": 11720
+    },
+    {
+      "epoch": 0.15449807864399642,
+      "grad_norm": 1.292722225189209,
+      "learning_rate": 9.599040037492968e-06,
+      "loss": 0.2133,
+      "step": 11740
+    },
+    {
+      "epoch": 0.15476127809654155,
+      "grad_norm": 2.0712890625,
+      "learning_rate": 9.597366007208095e-06,
+      "loss": 0.1996,
+      "step": 11760
+    },
+    {
+      "epoch": 0.1550244775490867,
+      "grad_norm": 1.482246994972229,
+      "learning_rate": 9.595688636207383e-06,
+      "loss": 0.1854,
+      "step": 11780
+    },
+    {
+      "epoch": 0.15528767700163185,
+      "grad_norm": 1.6078402996063232,
+      "learning_rate": 9.594007925709703e-06,
+      "loss": 0.1974,
+      "step": 11800
+    },
+    {
+      "epoch": 0.15555087645417698,
+      "grad_norm": 1.0145021677017212,
+      "learning_rate": 9.592323876936362e-06,
+      "loss": 0.2123,
+      "step": 11820
+    },
+    {
+      "epoch": 0.1558140759067221,
+      "grad_norm": 1.508529543876648,
+      "learning_rate": 9.59063649111109e-06,
+      "loss": 0.2234,
+      "step": 11840
+    },
+    {
+      "epoch": 0.15607727535926724,
+      "grad_norm": 1.6414767503738403,
+      "learning_rate": 9.588945769460036e-06,
+      "loss": 0.2115,
+      "step": 11860
+    },
+    {
+      "epoch": 0.1563404748118124,
+      "grad_norm": 1.176742434501648,
+      "learning_rate": 9.58725171321178e-06,
+      "loss": 0.2139,
+      "step": 11880
+    },
+    {
+      "epoch": 0.15660367426435753,
+      "grad_norm": 1.2484261989593506,
+      "learning_rate": 9.585554323597322e-06,
+      "loss": 0.2111,
+      "step": 11900
+    },
+    {
+      "epoch": 0.15686687371690267,
+      "grad_norm": 1.425335168838501,
+      "learning_rate": 9.583853601850087e-06,
+      "loss": 0.2001,
+      "step": 11920
+    },
+    {
+      "epoch": 0.1571300731694478,
+      "grad_norm": 1.1284537315368652,
+      "learning_rate": 9.582149549205916e-06,
+      "loss": 0.2204,
+      "step": 11940
+    },
+    {
+      "epoch": 0.15739327262199296,
+      "grad_norm": 1.3094009160995483,
+      "learning_rate": 9.580442166903074e-06,
+      "loss": 0.2052,
+      "step": 11960
+    },
+    {
+      "epoch": 0.1576564720745381,
+      "grad_norm": 1.3106313943862915,
+      "learning_rate": 9.578731456182249e-06,
+      "loss": 0.2108,
+      "step": 11980
+    },
+    {
+      "epoch": 0.15791967152708322,
+      "grad_norm": 1.2512283325195312,
+      "learning_rate": 9.577017418286538e-06,
+      "loss": 0.2173,
+      "step": 12000
+    },
+    {
+      "epoch": 0.15791967152708322,
+      "eval_loss": 0.2448178082704544,
+      "eval_runtime": 4.0971,
+      "eval_samples_per_second": 3.905,
+      "eval_steps_per_second": 1.953,
+      "step": 12000
+    },
+    {
+      "epoch": 0.15818287097962835,
+      "grad_norm": 2.1251821517944336,
+      "learning_rate": 9.575300054461467e-06,
+      "loss": 0.2112,
+      "step": 12020
+    },
+    {
+      "epoch": 0.1584460704321735,
+      "grad_norm": 1.386717677116394,
+      "learning_rate": 9.573579365954971e-06,
+      "loss": 0.2122,
+      "step": 12040
+    },
+    {
+      "epoch": 0.15870926988471865,
+      "grad_norm": 1.1348847150802612,
+      "learning_rate": 9.571855354017401e-06,
+      "loss": 0.2184,
+      "step": 12060
+    },
+    {
+      "epoch": 0.15897246933726378,
+      "grad_norm": 1.1104776859283447,
+      "learning_rate": 9.570128019901529e-06,
+      "loss": 0.2089,
+      "step": 12080
+    },
+    {
+      "epoch": 0.1592356687898089,
+      "grad_norm": 1.5507099628448486,
+      "learning_rate": 9.568397364862532e-06,
+      "loss": 0.2132,
+      "step": 12100
+    },
+    {
+      "epoch": 0.15949886824235407,
+      "grad_norm": 1.2343671321868896,
+      "learning_rate": 9.566663390158012e-06,
+      "loss": 0.2085,
+      "step": 12120
+    },
+    {
+      "epoch": 0.1597620676948992,
+      "grad_norm": 1.5279556512832642,
+      "learning_rate": 9.564926097047973e-06,
+      "loss": 0.2261,
+      "step": 12140
+    },
+    {
+      "epoch": 0.16002526714744433,
+      "grad_norm": 1.2676074504852295,
+      "learning_rate": 9.563185486794835e-06,
+      "loss": 0.2187,
+      "step": 12160
+    },
+    {
+      "epoch": 0.16028846659998947,
+      "grad_norm": 1.505919098854065,
+      "learning_rate": 9.561441560663427e-06,
+      "loss": 0.2003,
+      "step": 12180
+    },
+    {
+      "epoch": 0.1605516660525346,
+      "grad_norm": 1.590748906135559,
+      "learning_rate": 9.55969431992099e-06,
+      "loss": 0.1989,
+      "step": 12200
+    },
+    {
+      "epoch": 0.16081486550507976,
+      "grad_norm": 1.051192045211792,
+      "learning_rate": 9.557943765837167e-06,
+      "loss": 0.2221,
+      "step": 12220
+    },
+    {
+      "epoch": 0.1610780649576249,
+      "grad_norm": 1.2396160364151,
+      "learning_rate": 9.556189899684017e-06,
+      "loss": 0.2065,
+      "step": 12240
+    },
+    {
+      "epoch": 0.16134126441017002,
+      "grad_norm": 1.7275432348251343,
+      "learning_rate": 9.554432722736002e-06,
+      "loss": 0.1912,
+      "step": 12260
+    },
+    {
+      "epoch": 0.16160446386271515,
+      "grad_norm": 1.9184871912002563,
+      "learning_rate": 9.552672236269989e-06,
+      "loss": 0.2092,
+      "step": 12280
+    },
+    {
+      "epoch": 0.1618676633152603,
+      "grad_norm": 1.4065016508102417,
+      "learning_rate": 9.550908441565252e-06,
+      "loss": 0.2055,
+      "step": 12300
+    },
+    {
+      "epoch": 0.16213086276780544,
+      "grad_norm": 1.2429331541061401,
+      "learning_rate": 9.549141339903465e-06,
+      "loss": 0.2049,
+      "step": 12320
+    },
+    {
+      "epoch": 0.16239406222035058,
+      "grad_norm": 1.6071542501449585,
+      "learning_rate": 9.547370932568712e-06,
+      "loss": 0.2077,
+      "step": 12340
+    },
+    {
+      "epoch": 0.1626572616728957,
+      "grad_norm": 1.6860952377319336,
+      "learning_rate": 9.545597220847469e-06,
+      "loss": 0.2093,
+      "step": 12360
+    },
+    {
+      "epoch": 0.16292046112544087,
+      "grad_norm": 1.828596830368042,
+      "learning_rate": 9.543820206028623e-06,
+      "loss": 0.2057,
+      "step": 12380
+    },
+    {
+      "epoch": 0.163183660577986,
+      "grad_norm": 1.8210543394088745,
+      "learning_rate": 9.542039889403458e-06,
+      "loss": 0.1922,
+      "step": 12400
+    },
+    {
+      "epoch": 0.16344686003053113,
+      "grad_norm": 1.3486182689666748,
+      "learning_rate": 9.540256272265653e-06,
+      "loss": 0.2089,
+      "step": 12420
+    },
+    {
+      "epoch": 0.16371005948307626,
+      "grad_norm": 1.4413331747055054,
+      "learning_rate": 9.53846935591129e-06,
+      "loss": 0.2145,
+      "step": 12440
+    },
+    {
+      "epoch": 0.16397325893562142,
+      "grad_norm": 1.6699386835098267,
+      "learning_rate": 9.536679141638851e-06,
+      "loss": 0.2194,
+      "step": 12460
+    },
+    {
+      "epoch": 0.16423645838816656,
+      "grad_norm": 1.3608828783035278,
+      "learning_rate": 9.534885630749206e-06,
+      "loss": 0.2164,
+      "step": 12480
+    },
+    {
+      "epoch": 0.1644996578407117,
+      "grad_norm": 1.0602197647094727,
+      "learning_rate": 9.533088824545626e-06,
+      "loss": 0.2242,
+      "step": 12500
+    },
+    {
+      "epoch": 0.1644996578407117,
+      "eval_loss": 0.2395314872264862,
+      "eval_runtime": 3.5914,
+      "eval_samples_per_second": 4.455,
+      "eval_steps_per_second": 2.228,
+      "step": 12500
+    },
+    {
+      "epoch": 0.16476285729325682,
+      "grad_norm": 1.2981318235397339,
+      "learning_rate": 9.531288724333777e-06,
+      "loss": 0.2128,
+      "step": 12520
+    },
+    {
+      "epoch": 0.16502605674580198,
+      "grad_norm": 1.3617362976074219,
+      "learning_rate": 9.52948533142172e-06,
+      "loss": 0.2065,
+      "step": 12540
+    },
+    {
+      "epoch": 0.1652892561983471,
+      "grad_norm": 1.4593777656555176,
+      "learning_rate": 9.5276786471199e-06,
+      "loss": 0.194,
+      "step": 12560
+    },
+    {
+      "epoch": 0.16555245565089224,
+      "grad_norm": 1.357919454574585,
+      "learning_rate": 9.525868672741164e-06,
+      "loss": 0.2078,
+      "step": 12580
+    },
+    {
+      "epoch": 0.16581565510343738,
+      "grad_norm": 1.3947018384933472,
+      "learning_rate": 9.524055409600745e-06,
+      "loss": 0.2003,
+      "step": 12600
+    },
+    {
+      "epoch": 0.16607885455598254,
+      "grad_norm": 1.4510191679000854,
+      "learning_rate": 9.522238859016264e-06,
+      "loss": 0.2065,
+      "step": 12620
+    },
+    {
+      "epoch": 0.16634205400852767,
+      "grad_norm": 1.6841310262680054,
+      "learning_rate": 9.520419022307737e-06,
+      "loss": 0.1876,
+      "step": 12640
+    },
+    {
+      "epoch": 0.1666052534610728,
+      "grad_norm": 1.626165509223938,
+      "learning_rate": 9.518595900797563e-06,
+      "loss": 0.2113,
+      "step": 12660
+    },
+    {
+      "epoch": 0.16686845291361793,
+      "grad_norm": 1.7602510452270508,
+      "learning_rate": 9.516769495810526e-06,
+      "loss": 0.213,
+      "step": 12680
+    },
+    {
+      "epoch": 0.1671316523661631,
+      "grad_norm": 1.43569016456604,
+      "learning_rate": 9.514939808673803e-06,
+      "loss": 0.2143,
+      "step": 12700
+    },
+    {
+      "epoch": 0.16739485181870822,
+      "grad_norm": 1.4139032363891602,
+      "learning_rate": 9.51310684071695e-06,
+      "loss": 0.2168,
+      "step": 12720
+    },
+    {
+      "epoch": 0.16765805127125336,
+      "grad_norm": 1.198973298072815,
+      "learning_rate": 9.511270593271908e-06,
+      "loss": 0.2088,
+      "step": 12740
+    },
+    {
+      "epoch": 0.1679212507237985,
+      "grad_norm": 1.5792280435562134,
+      "learning_rate": 9.509431067673005e-06,
+      "loss": 0.2314,
+      "step": 12760
+    },
+    {
+      "epoch": 0.16818445017634362,
+      "grad_norm": 1.7921568155288696,
+      "learning_rate": 9.507588265256947e-06,
+      "loss": 0.1956,
+      "step": 12780
+    },
+    {
+      "epoch": 0.16844764962888878,
+      "grad_norm": 1.0051674842834473,
+      "learning_rate": 9.505742187362822e-06,
+      "loss": 0.2214,
+      "step": 12800
+    },
+    {
+      "epoch": 0.1687108490814339,
+      "grad_norm": 1.352509617805481,
+      "learning_rate": 9.503892835332099e-06,
+      "loss": 0.2151,
+      "step": 12820
+    },
+    {
+      "epoch": 0.16897404853397904,
+      "grad_norm": 1.3784739971160889,
+      "learning_rate": 9.502040210508624e-06,
+      "loss": 0.2151,
+      "step": 12840
+    },
+    {
+      "epoch": 0.16923724798652418,
+      "grad_norm": 1.2261897325515747,
+      "learning_rate": 9.500184314238622e-06,
+      "loss": 0.2319,
+      "step": 12860
+    },
+    {
+      "epoch": 0.16950044743906933,
+      "grad_norm": 1.9484477043151855,
+      "learning_rate": 9.498325147870702e-06,
+      "loss": 0.1908,
+      "step": 12880
+    },
+    {
+      "epoch": 0.16976364689161447,
+      "grad_norm": 1.3108311891555786,
+      "learning_rate": 9.496462712755839e-06,
+      "loss": 0.2008,
+      "step": 12900
+    },
+    {
+      "epoch": 0.1700268463441596,
+      "grad_norm": 1.2683104276657104,
+      "learning_rate": 9.494597010247387e-06,
+      "loss": 0.1873,
+      "step": 12920
+    },
+    {
+      "epoch": 0.17029004579670473,
+      "grad_norm": 1.1195465326309204,
+      "learning_rate": 9.492728041701077e-06,
+      "loss": 0.2042,
+      "step": 12940
+    },
+    {
+      "epoch": 0.1705532452492499,
+      "grad_norm": 2.3313589096069336,
+      "learning_rate": 9.490855808475007e-06,
+      "loss": 0.2142,
+      "step": 12960
+    },
+    {
+      "epoch": 0.17081644470179502,
+      "grad_norm": 1.4149857759475708,
+      "learning_rate": 9.488980311929658e-06,
+      "loss": 0.2048,
+      "step": 12980
+    },
+    {
+      "epoch": 0.17107964415434015,
+      "grad_norm": 1.749277949333191,
+      "learning_rate": 9.487101553427872e-06,
+      "loss": 0.1965,
+      "step": 13000
+    },
+    {
+      "epoch": 0.17107964415434015,
+      "eval_loss": 0.23680630326271057,
+      "eval_runtime": 4.4792,
+      "eval_samples_per_second": 3.572,
+      "eval_steps_per_second": 1.786,
+      "step": 13000
+    },
+    {
+      "epoch": 0.1713428436068853,
+      "grad_norm": 1.561540961265564,
+      "learning_rate": 9.485219534334867e-06,
+      "loss": 0.1977,
+      "step": 13020
+    },
+    {
+      "epoch": 0.17160604305943045,
+      "grad_norm": 1.6069048643112183,
+      "learning_rate": 9.483334256018225e-06,
+      "loss": 0.2044,
+      "step": 13040
+    },
+    {
+      "epoch": 0.17186924251197558,
+      "grad_norm": 1.4624124765396118,
+      "learning_rate": 9.481445719847901e-06,
+      "loss": 0.1945,
+      "step": 13060
+    },
+    {
+      "epoch": 0.1721324419645207,
+      "grad_norm": 1.3036525249481201,
+      "learning_rate": 9.47955392719622e-06,
+      "loss": 0.2383,
+      "step": 13080
+    },
+    {
+      "epoch": 0.17239564141706584,
+      "grad_norm": 1.2367247343063354,
+      "learning_rate": 9.477658879437866e-06,
+      "loss": 0.2042,
+      "step": 13100
+    },
+    {
+      "epoch": 0.172658840869611,
+      "grad_norm": 1.4068106412887573,
+      "learning_rate": 9.475760577949893e-06,
+      "loss": 0.1954,
+      "step": 13120
+    },
+    {
+      "epoch": 0.17292204032215613,
+      "grad_norm": 1.277147650718689,
+      "learning_rate": 9.473859024111715e-06,
+      "loss": 0.2178,
+      "step": 13140
+    },
+    {
+      "epoch": 0.17318523977470127,
+      "grad_norm": 1.5708671808242798,
+      "learning_rate": 9.471954219305118e-06,
+      "loss": 0.2019,
+      "step": 13160
+    },
+    {
+      "epoch": 0.1734484392272464,
+      "grad_norm": 1.1117615699768066,
+      "learning_rate": 9.470046164914242e-06,
+      "loss": 0.2032,
+      "step": 13180
+    },
+    {
+      "epoch": 0.17371163867979156,
+      "grad_norm": 1.5927276611328125,
+      "learning_rate": 9.468134862325591e-06,
+      "loss": 0.2065,
+      "step": 13200
+    },
+    {
+      "epoch": 0.1739748381323367,
+      "grad_norm": 1.2429004907608032,
+      "learning_rate": 9.46622031292803e-06,
+      "loss": 0.1904,
+      "step": 13220
+    },
+    {
+      "epoch": 0.17423803758488182,
+      "grad_norm": 1.704650640487671,
+      "learning_rate": 9.464302518112783e-06,
+      "loss": 0.2043,
+      "step": 13240
+    },
+    {
+      "epoch": 0.17450123703742695,
+      "grad_norm": 1.319659948348999,
+      "learning_rate": 9.462381479273433e-06,
+      "loss": 0.2403,
+      "step": 13260
+    },
+    {
+      "epoch": 0.1747644364899721,
+      "grad_norm": 1.4926093816757202,
+      "learning_rate": 9.46045719780592e-06,
+      "loss": 0.1967,
+      "step": 13280
+    },
+    {
+      "epoch": 0.17502763594251725,
+      "grad_norm": 1.6184725761413574,
+      "learning_rate": 9.458529675108536e-06,
+      "loss": 0.1917,
+      "step": 13300
+    },
+    {
+      "epoch": 0.17529083539506238,
+      "grad_norm": 2.4485104084014893,
+      "learning_rate": 9.456598912581935e-06,
+      "loss": 0.1858,
+      "step": 13320
+    },
+    {
+      "epoch": 0.1755540348476075,
+      "grad_norm": 1.4941998720169067,
+      "learning_rate": 9.454664911629122e-06,
+      "loss": 0.204,
+      "step": 13340
+    },
+    {
+      "epoch": 0.17581723430015267,
+      "grad_norm": 2.0899596214294434,
+      "learning_rate": 9.452727673655454e-06,
+      "loss": 0.2049,
+      "step": 13360
+    },
+    {
+      "epoch": 0.1760804337526978,
+      "grad_norm": 1.8336759805679321,
+      "learning_rate": 9.450787200068647e-06,
+      "loss": 0.1966,
+      "step": 13380
+    },
+    {
+      "epoch": 0.17634363320524293,
+      "grad_norm": 1.3282217979431152,
+      "learning_rate": 9.448843492278757e-06,
+      "loss": 0.2196,
+      "step": 13400
+    },
+    {
+      "epoch": 0.17660683265778807,
+      "grad_norm": 2.1068334579467773,
+      "learning_rate": 9.4468965516982e-06,
+      "loss": 0.1953,
+      "step": 13420
+    },
+    {
+      "epoch": 0.1768700321103332,
+      "grad_norm": 1.4248470067977905,
+      "learning_rate": 9.444946379741732e-06,
+      "loss": 0.2005,
+      "step": 13440
+    },
+    {
+      "epoch": 0.17713323156287836,
+      "grad_norm": 1.1936007738113403,
+      "learning_rate": 9.44299297782647e-06,
+      "loss": 0.2235,
+      "step": 13460
+    },
+    {
+      "epoch": 0.1773964310154235,
+      "grad_norm": 1.6337552070617676,
+      "learning_rate": 9.441036347371864e-06,
+      "loss": 0.1965,
+      "step": 13480
+    },
+    {
+      "epoch": 0.17765963046796862,
+      "grad_norm": 1.7119965553283691,
+      "learning_rate": 9.439076489799719e-06,
+      "loss": 0.2063,
+      "step": 13500
+    },
+    {
+      "epoch": 0.17765963046796862,
+      "eval_loss": 0.23400452733039856,
+      "eval_runtime": 4.05,
+      "eval_samples_per_second": 3.951,
+      "eval_steps_per_second": 1.975,
+      "step": 13500
+    },
+    {
+      "epoch": 0.17792282992051375,
+      "grad_norm": 1.8556243181228638,
+      "learning_rate": 9.437113406534182e-06,
+      "loss": 0.2027,
+      "step": 13520
+    },
+    {
+      "epoch": 0.1781860293730589,
+      "grad_norm": 1.2260537147521973,
+      "learning_rate": 9.435147099001745e-06,
+      "loss": 0.208,
+      "step": 13540
+    },
+    {
+      "epoch": 0.17844922882560404,
+      "grad_norm": 1.0981372594833374,
+      "learning_rate": 9.433177568631243e-06,
+      "loss": 0.2155,
+      "step": 13560
+    },
+    {
+      "epoch": 0.17871242827814918,
+      "grad_norm": 1.6278541088104248,
+      "learning_rate": 9.43120481685385e-06,
+      "loss": 0.1866,
+      "step": 13580
+    },
+    {
+      "epoch": 0.1789756277306943,
+      "grad_norm": 1.5484399795532227,
+      "learning_rate": 9.429228845103087e-06,
+      "loss": 0.1974,
+      "step": 13600
+    },
+    {
+      "epoch": 0.17923882718323947,
+      "grad_norm": 1.0902371406555176,
+      "learning_rate": 9.427249654814806e-06,
+      "loss": 0.2259,
+      "step": 13620
+    },
+    {
+      "epoch": 0.1795020266357846,
+      "grad_norm": 1.601146936416626,
+      "learning_rate": 9.425267247427207e-06,
+      "loss": 0.2001,
+      "step": 13640
+    },
+    {
+      "epoch": 0.17976522608832973,
+      "grad_norm": 1.4625359773635864,
+      "learning_rate": 9.423281624380821e-06,
+      "loss": 0.2007,
+      "step": 13660
+    },
+    {
+      "epoch": 0.18002842554087486,
+      "grad_norm": 1.6898537874221802,
+      "learning_rate": 9.42129278711852e-06,
+      "loss": 0.2067,
+      "step": 13680
+    },
+    {
+      "epoch": 0.18029162499342002,
+      "grad_norm": 1.3530702590942383,
+      "learning_rate": 9.419300737085509e-06,
+      "loss": 0.2157,
+      "step": 13700
+    },
+    {
+      "epoch": 0.18055482444596516,
+      "grad_norm": 1.044854760169983,
+      "learning_rate": 9.417305475729325e-06,
+      "loss": 0.2154,
+      "step": 13720
+    },
+    {
+      "epoch": 0.1808180238985103,
+      "grad_norm": 2.1911394596099854,
+      "learning_rate": 9.41530700449985e-06,
+      "loss": 0.2244,
+      "step": 13740
+    },
+    {
+      "epoch": 0.18108122335105542,
+      "grad_norm": 1.5178133249282837,
+      "learning_rate": 9.413305324849282e-06,
+      "loss": 0.1977,
+      "step": 13760
+    },
+    {
+      "epoch": 0.18134442280360058,
+      "grad_norm": 1.314210057258606,
+      "learning_rate": 9.411300438232164e-06,
+      "loss": 0.1998,
+      "step": 13780
+    },
+    {
+      "epoch": 0.1816076222561457,
+      "grad_norm": 1.2863247394561768,
+      "learning_rate": 9.409292346105365e-06,
+      "loss": 0.2169,
+      "step": 13800
+    },
+    {
+      "epoch": 0.18187082170869084,
+      "grad_norm": 1.315562129020691,
+      "learning_rate": 9.407281049928078e-06,
+      "loss": 0.2,
+      "step": 13820
+    },
+    {
+      "epoch": 0.18213402116123598,
+      "grad_norm": 1.1411410570144653,
+      "learning_rate": 9.405266551161831e-06,
+      "loss": 0.2246,
+      "step": 13840
+    },
+    {
+      "epoch": 0.18239722061378114,
+      "grad_norm": 1.1800204515457153,
+      "learning_rate": 9.403248851270477e-06,
+      "loss": 0.203,
+      "step": 13860
+    },
+    {
+      "epoch": 0.18266042006632627,
+      "grad_norm": 1.472043514251709,
+      "learning_rate": 9.401227951720195e-06,
+      "loss": 0.2174,
+      "step": 13880
+    },
+    {
+      "epoch": 0.1829236195188714,
+      "grad_norm": 1.5994408130645752,
+      "learning_rate": 9.399203853979491e-06,
+      "loss": 0.2138,
+      "step": 13900
+    },
+    {
+      "epoch": 0.18318681897141653,
+      "grad_norm": 1.5901877880096436,
+      "learning_rate": 9.39717655951919e-06,
+      "loss": 0.1965,
+      "step": 13920
+    },
+    {
+      "epoch": 0.1834500184239617,
+      "grad_norm": 1.9576339721679688,
+      "learning_rate": 9.395146069812447e-06,
+      "loss": 0.2139,
+      "step": 13940
+    },
+    {
+      "epoch": 0.18371321787650682,
+      "grad_norm": 1.655224084854126,
+      "learning_rate": 9.39311238633473e-06,
+      "loss": 0.2037,
+      "step": 13960
+    },
+    {
+      "epoch": 0.18397641732905196,
+      "grad_norm": 1.5874608755111694,
+      "learning_rate": 9.391075510563836e-06,
+      "loss": 0.1924,
+      "step": 13980
+    },
+    {
+      "epoch": 0.1842396167815971,
+      "grad_norm": 1.0008459091186523,
+      "learning_rate": 9.389035443979878e-06,
+      "loss": 0.2005,
+      "step": 14000
+    },
+    {
+      "epoch": 0.1842396167815971,
+      "eval_loss": 0.23292672634124756,
+      "eval_runtime": 4.28,
+      "eval_samples_per_second": 3.738,
+      "eval_steps_per_second": 1.869,
+      "step": 14000
+    },
+    {
+      "epoch": 0.18450281623414222,
+      "grad_norm": 1.5699293613433838,
+      "learning_rate": 9.386992188065289e-06,
+      "loss": 0.2019,
+      "step": 14020
+    },
+    {
+      "epoch": 0.18476601568668738,
+      "grad_norm": 1.601386308670044,
+      "learning_rate": 9.384945744304817e-06,
+      "loss": 0.1932,
+      "step": 14040
+    },
+    {
+      "epoch": 0.1850292151392325,
+      "grad_norm": 1.118137240409851,
+      "learning_rate": 9.382896114185527e-06,
+      "loss": 0.1794,
+      "step": 14060
+    },
+    {
+      "epoch": 0.18529241459177764,
+      "grad_norm": 1.3343223333358765,
+      "learning_rate": 9.380843299196802e-06,
+      "loss": 0.1962,
+      "step": 14080
+    },
+    {
+      "epoch": 0.18555561404432278,
+      "grad_norm": 1.1525280475616455,
+      "learning_rate": 9.378787300830338e-06,
+      "loss": 0.2118,
+      "step": 14100
+    },
+    {
+      "epoch": 0.18581881349686794,
+      "grad_norm": 1.6061981916427612,
+      "learning_rate": 9.376728120580144e-06,
+      "loss": 0.1996,
+      "step": 14120
+    },
+    {
+      "epoch": 0.18608201294941307,
+      "grad_norm": 1.2430217266082764,
+      "learning_rate": 9.37466575994254e-06,
+      "loss": 0.1926,
+      "step": 14140
+    },
+    {
+      "epoch": 0.1863452124019582,
+      "grad_norm": 1.5858440399169922,
+      "learning_rate": 9.372600220416161e-06,
+      "loss": 0.2182,
+      "step": 14160
+    },
+    {
+      "epoch": 0.18660841185450333,
+      "grad_norm": 2.3179852962493896,
+      "learning_rate": 9.370531503501947e-06,
+      "loss": 0.2043,
+      "step": 14180
+    },
+    {
+      "epoch": 0.1868716113070485,
+      "grad_norm": 1.2648091316223145,
+      "learning_rate": 9.368459610703148e-06,
+      "loss": 0.1959,
+      "step": 14200
+    },
+    {
+      "epoch": 0.18713481075959362,
+      "grad_norm": 1.5083450078964233,
+      "learning_rate": 9.366384543525327e-06,
+      "loss": 0.2234,
+      "step": 14220
+    },
+    {
+      "epoch": 0.18739801021213875,
+      "grad_norm": 1.554119348526001,
+      "learning_rate": 9.364306303476344e-06,
+      "loss": 0.2076,
+      "step": 14240
+    },
+    {
+      "epoch": 0.1876612096646839,
+      "grad_norm": 1.1560771465301514,
+      "learning_rate": 9.362224892066376e-06,
+      "loss": 0.2001,
+      "step": 14260
+    },
+    {
+      "epoch": 0.18792440911722905,
+      "grad_norm": 1.3132483959197998,
+      "learning_rate": 9.360140310807893e-06,
+      "loss": 0.2112,
+      "step": 14280
+    },
+    {
+      "epoch": 0.18818760856977418,
+      "grad_norm": 1.1594722270965576,
+      "learning_rate": 9.358052561215679e-06,
+      "loss": 0.1968,
+      "step": 14300
+    },
+    {
+      "epoch": 0.1884508080223193,
+      "grad_norm": 1.2698969841003418,
+      "learning_rate": 9.35596164480681e-06,
+      "loss": 0.213,
+      "step": 14320
+    },
+    {
+      "epoch": 0.18871400747486444,
+      "grad_norm": 1.0864530801773071,
+      "learning_rate": 9.353867563100673e-06,
+      "loss": 0.206,
+      "step": 14340
+    },
+    {
+      "epoch": 0.1889772069274096,
+      "grad_norm": 1.5392730236053467,
+      "learning_rate": 9.35177031761895e-06,
+      "loss": 0.2146,
+      "step": 14360
+    },
+    {
+      "epoch": 0.18924040637995473,
+      "grad_norm": 1.1666347980499268,
+      "learning_rate": 9.349669909885619e-06,
+      "loss": 0.2041,
+      "step": 14380
+    },
+    {
+      "epoch": 0.18950360583249987,
+      "grad_norm": 1.440243124961853,
+      "learning_rate": 9.347566341426965e-06,
+      "loss": 0.2011,
+      "step": 14400
+    },
+    {
+      "epoch": 0.189766805285045,
+      "grad_norm": 1.2161346673965454,
+      "learning_rate": 9.345459613771561e-06,
+      "loss": 0.1964,
+      "step": 14420
+    },
+    {
+      "epoch": 0.19003000473759016,
+      "grad_norm": 1.4280766248703003,
+      "learning_rate": 9.343349728450278e-06,
+      "loss": 0.219,
+      "step": 14440
+    },
+    {
+      "epoch": 0.1902932041901353,
+      "grad_norm": 1.570294737815857,
+      "learning_rate": 9.341236686996285e-06,
+      "loss": 0.1992,
+      "step": 14460
+    },
+    {
+      "epoch": 0.19055640364268042,
+      "grad_norm": 1.4833028316497803,
+      "learning_rate": 9.33912049094504e-06,
+      "loss": 0.2123,
+      "step": 14480
+    },
+    {
+      "epoch": 0.19081960309522555,
+      "grad_norm": 2.0131571292877197,
+      "learning_rate": 9.337001141834296e-06,
+      "loss": 0.2259,
+      "step": 14500
+    },
+    {
+      "epoch": 0.19081960309522555,
+      "eval_loss": 0.236808180809021,
+      "eval_runtime": 4.1567,
+      "eval_samples_per_second": 3.849,
+      "eval_steps_per_second": 1.925,
+      "step": 14500
+    },
+    {
+      "epoch": 0.1910828025477707,
+      "grad_norm": 1.4779287576675415,
+      "learning_rate": 9.334878641204098e-06,
+      "loss": 0.2051,
+      "step": 14520
+    },
+    {
+      "epoch": 0.19134600200031585,
+      "grad_norm": 1.379014492034912,
+      "learning_rate": 9.332752990596773e-06,
+      "loss": 0.1926,
+      "step": 14540
+    },
+    {
+      "epoch": 0.19160920145286098,
+      "grad_norm": 1.4269139766693115,
+      "learning_rate": 9.330624191556952e-06,
+      "loss": 0.2032,
+      "step": 14560
+    },
+    {
+      "epoch": 0.1918724009054061,
+      "grad_norm": 1.4228744506835938,
+      "learning_rate": 9.328492245631542e-06,
+      "loss": 0.2164,
+      "step": 14580
+    },
+    {
+      "epoch": 0.19213560035795124,
+      "grad_norm": 2.1166574954986572,
+      "learning_rate": 9.326357154369739e-06,
+      "loss": 0.2097,
+      "step": 14600
+    },
+    {
+      "epoch": 0.1923987998104964,
+      "grad_norm": 1.4721271991729736,
+      "learning_rate": 9.324218919323024e-06,
+      "loss": 0.203,
+      "step": 14620
+    },
+    {
+      "epoch": 0.19266199926304153,
+      "grad_norm": 1.4993586540222168,
+      "learning_rate": 9.322077542045167e-06,
+      "loss": 0.199,
+      "step": 14640
+    },
+    {
+      "epoch": 0.19292519871558667,
+      "grad_norm": 1.2647351026535034,
+      "learning_rate": 9.319933024092216e-06,
+      "loss": 0.2237,
+      "step": 14660
+    },
+    {
+      "epoch": 0.1931883981681318,
+      "grad_norm": 1.5198867321014404,
+      "learning_rate": 9.317785367022506e-06,
+      "loss": 0.2028,
+      "step": 14680
+    },
+    {
+      "epoch": 0.19345159762067696,
+      "grad_norm": 1.3967750072479248,
+      "learning_rate": 9.315634572396652e-06,
+      "loss": 0.1864,
+      "step": 14700
+    },
+    {
+      "epoch": 0.1937147970732221,
+      "grad_norm": 1.1446253061294556,
+      "learning_rate": 9.313480641777542e-06,
+      "loss": 0.2188,
+      "step": 14720
+    },
+    {
+      "epoch": 0.19397799652576722,
+      "grad_norm": 1.4299098253250122,
+      "learning_rate": 9.311323576730353e-06,
+      "loss": 0.1973,
+      "step": 14740
+    },
+    {
+      "epoch": 0.19424119597831235,
+      "grad_norm": 1.2577261924743652,
+      "learning_rate": 9.309163378822532e-06,
+      "loss": 0.2229,
+      "step": 14760
+    },
+    {
+      "epoch": 0.1945043954308575,
+      "grad_norm": 1.5158898830413818,
+      "learning_rate": 9.307000049623807e-06,
+      "loss": 0.2006,
+      "step": 14780
+    },
+    {
+      "epoch": 0.19476759488340264,
+      "grad_norm": 1.2390705347061157,
+      "learning_rate": 9.304833590706181e-06,
+      "loss": 0.1852,
+      "step": 14800
+    },
+    {
+      "epoch": 0.19503079433594778,
+      "grad_norm": 1.6396849155426025,
+      "learning_rate": 9.302664003643929e-06,
+      "loss": 0.191,
+      "step": 14820
+    },
+    {
+      "epoch": 0.1952939937884929,
+      "grad_norm": 1.2442011833190918,
+      "learning_rate": 9.3004912900136e-06,
+      "loss": 0.1999,
+      "step": 14840
+    },
+    {
+      "epoch": 0.19555719324103807,
+      "grad_norm": 0.9597941637039185,
+      "learning_rate": 9.298315451394015e-06,
+      "loss": 0.194,
+      "step": 14860
+    },
+    {
+      "epoch": 0.1958203926935832,
+      "grad_norm": 1.4320135116577148,
+      "learning_rate": 9.296136489366266e-06,
+      "loss": 0.2054,
+      "step": 14880
+    },
+    {
+      "epoch": 0.19608359214612833,
+      "grad_norm": 1.7578682899475098,
+      "learning_rate": 9.293954405513717e-06,
+      "loss": 0.2129,
+      "step": 14900
+    },
+    {
+      "epoch": 0.19634679159867346,
+      "grad_norm": 1.3280483484268188,
+      "learning_rate": 9.291769201421994e-06,
+      "loss": 0.2124,
+      "step": 14920
+    },
+    {
+      "epoch": 0.19660999105121862,
+      "grad_norm": 1.6327404975891113,
+      "learning_rate": 9.289580878678996e-06,
+      "loss": 0.1863,
+      "step": 14940
+    },
+    {
+      "epoch": 0.19687319050376376,
+      "grad_norm": 1.397274136543274,
+      "learning_rate": 9.28738943887489e-06,
+      "loss": 0.2071,
+      "step": 14960
+    },
+    {
+      "epoch": 0.1971363899563089,
+      "grad_norm": 1.6567870378494263,
+      "learning_rate": 9.2851948836021e-06,
+      "loss": 0.2153,
+      "step": 14980
+    },
+    {
+      "epoch": 0.19739958940885402,
+      "grad_norm": 1.4747213125228882,
+      "learning_rate": 9.282997214455321e-06,
+      "loss": 0.1982,
+      "step": 15000
+    },
+    {
+      "epoch": 0.19739958940885402,
+      "eval_loss": 0.23930305242538452,
+      "eval_runtime": 4.3202,
+      "eval_samples_per_second": 3.704,
+      "eval_steps_per_second": 1.852,
+      "step": 15000
+    },
+    {
+      "epoch": 0.19766278886139918,
+      "grad_norm": 1.9889999628067017,
+      "learning_rate": 9.28079643303151e-06,
+      "loss": 0.1993,
+      "step": 15020
+    },
+    {
+      "epoch": 0.1979259883139443,
+      "grad_norm": 1.5881109237670898,
+      "learning_rate": 9.27859254092988e-06,
+      "loss": 0.2082,
+      "step": 15040
+    },
+    {
+      "epoch": 0.19818918776648944,
+      "grad_norm": 1.79412043094635,
+      "learning_rate": 9.27638553975191e-06,
+      "loss": 0.1883,
+      "step": 15060
+    },
+    {
+      "epoch": 0.19845238721903458,
+      "grad_norm": 1.4508376121520996,
+      "learning_rate": 9.274175431101337e-06,
+      "loss": 0.2175,
+      "step": 15080
+    },
+    {
+      "epoch": 0.19871558667157974,
+      "grad_norm": 0.9127804040908813,
+      "learning_rate": 9.271962216584156e-06,
+      "loss": 0.2015,
+      "step": 15100
+    },
+    {
+      "epoch": 0.19897878612412487,
+      "grad_norm": 1.6234700679779053,
+      "learning_rate": 9.269745897808617e-06,
+      "loss": 0.1909,
+      "step": 15120
+    },
+    {
+      "epoch": 0.19924198557667,
+      "grad_norm": 1.3188230991363525,
+      "learning_rate": 9.267526476385228e-06,
+      "loss": 0.1943,
+      "step": 15140
+    },
+    {
+      "epoch": 0.19950518502921513,
+      "grad_norm": 1.1227322816848755,
+      "learning_rate": 9.265303953926751e-06,
+      "loss": 0.2048,
+      "step": 15160
+    },
+    {
+      "epoch": 0.1997683844817603,
+      "grad_norm": 1.3703421354293823,
+      "learning_rate": 9.2630783320482e-06,
+      "loss": 0.2187,
+      "step": 15180
+    },
+    {
+      "epoch": 0.20003158393430542,
+      "grad_norm": 1.3907465934753418,
+      "learning_rate": 9.260849612366845e-06,
+      "loss": 0.2152,
+      "step": 15200
+    },
+    {
+      "epoch": 0.20029478338685056,
+      "grad_norm": 1.3484028577804565,
+      "learning_rate": 9.258617796502204e-06,
+      "loss": 0.2001,
+      "step": 15220
+    },
+    {
+      "epoch": 0.2005579828393957,
+      "grad_norm": 1.5996466875076294,
+      "learning_rate": 9.256382886076042e-06,
+      "loss": 0.1992,
+      "step": 15240
+    },
+    {
+      "epoch": 0.20082118229194082,
+      "grad_norm": 1.4887921810150146,
+      "learning_rate": 9.25414488271238e-06,
+      "loss": 0.2009,
+      "step": 15260
+    },
+    {
+      "epoch": 0.20108438174448598,
+      "grad_norm": 1.9896572828292847,
+      "learning_rate": 9.251903788037479e-06,
+      "loss": 0.2047,
+      "step": 15280
+    },
+    {
+      "epoch": 0.2013475811970311,
+      "grad_norm": 1.977705478668213,
+      "learning_rate": 9.249659603679853e-06,
+      "loss": 0.2065,
+      "step": 15300
+    },
+    {
+      "epoch": 0.20161078064957624,
+      "grad_norm": 1.6992383003234863,
+      "learning_rate": 9.247412331270256e-06,
+      "loss": 0.2019,
+      "step": 15320
+    },
+    {
+      "epoch": 0.20187398010212138,
+      "grad_norm": 2.001516342163086,
+      "learning_rate": 9.24516197244169e-06,
+      "loss": 0.1871,
+      "step": 15340
+    },
+    {
+      "epoch": 0.20213717955466654,
+      "grad_norm": 1.7694910764694214,
+      "learning_rate": 9.242908528829396e-06,
+      "loss": 0.1993,
+      "step": 15360
+    },
+    {
+      "epoch": 0.20240037900721167,
+      "grad_norm": 1.2791032791137695,
+      "learning_rate": 9.240652002070856e-06,
+      "loss": 0.2024,
+      "step": 15380
+    },
+    {
+      "epoch": 0.2026635784597568,
+      "grad_norm": 1.4787293672561646,
+      "learning_rate": 9.2383923938058e-06,
+      "loss": 0.1987,
+      "step": 15400
+    },
+    {
+      "epoch": 0.20292677791230193,
+      "grad_norm": 1.2745243310928345,
+      "learning_rate": 9.236129705676188e-06,
+      "loss": 0.1932,
+      "step": 15420
+    },
+    {
+      "epoch": 0.2031899773648471,
+      "grad_norm": 1.472952961921692,
+      "learning_rate": 9.233863939326225e-06,
+      "loss": 0.1931,
+      "step": 15440
+    },
+    {
+      "epoch": 0.20345317681739222,
+      "grad_norm": 1.7805614471435547,
+      "learning_rate": 9.231595096402347e-06,
+      "loss": 0.2034,
+      "step": 15460
+    },
+    {
+      "epoch": 0.20371637626993735,
+      "grad_norm": 1.1107107400894165,
+      "learning_rate": 9.22932317855323e-06,
+      "loss": 0.1904,
+      "step": 15480
+    },
+    {
+      "epoch": 0.2039795757224825,
+      "grad_norm": 1.1010551452636719,
+      "learning_rate": 9.227048187429781e-06,
+      "loss": 0.2056,
+      "step": 15500
+    },
+    {
+      "epoch": 0.2039795757224825,
+      "eval_loss": 0.23426011204719543,
+      "eval_runtime": 4.4676,
+      "eval_samples_per_second": 3.581,
+      "eval_steps_per_second": 1.791,
+      "step": 15500
+    },
+    {
+      "epoch": 0.20424277517502765,
+      "grad_norm": 1.5277096033096313,
+      "learning_rate": 9.224770124685145e-06,
+      "loss": 0.2145,
+      "step": 15520
+    },
+    {
+      "epoch": 0.20450597462757278,
+      "grad_norm": 1.8335736989974976,
+      "learning_rate": 9.222488991974693e-06,
+      "loss": 0.1824,
+      "step": 15540
+    },
+    {
+      "epoch": 0.2047691740801179,
+      "grad_norm": 1.496925950050354,
+      "learning_rate": 9.220204790956032e-06,
+      "loss": 0.2199,
+      "step": 15560
+    },
+    {
+      "epoch": 0.20503237353266304,
+      "grad_norm": 1.2986793518066406,
+      "learning_rate": 9.217917523288995e-06,
+      "loss": 0.2113,
+      "step": 15580
+    },
+    {
+      "epoch": 0.2052955729852082,
+      "grad_norm": 1.6191668510437012,
+      "learning_rate": 9.215627190635645e-06,
+      "loss": 0.2215,
+      "step": 15600
+    },
+    {
+      "epoch": 0.20555877243775333,
+      "grad_norm": 1.4210251569747925,
+      "learning_rate": 9.213333794660272e-06,
+      "loss": 0.2092,
+      "step": 15620
+    },
+    {
+      "epoch": 0.20582197189029847,
+      "grad_norm": 1.502590298652649,
+      "learning_rate": 9.211037337029394e-06,
+      "loss": 0.1835,
+      "step": 15640
+    },
+    {
+      "epoch": 0.2060851713428436,
+      "grad_norm": 1.8053871393203735,
+      "learning_rate": 9.20873781941175e-06,
+      "loss": 0.207,
+      "step": 15660
+    },
+    {
+      "epoch": 0.20634837079538876,
+      "grad_norm": 1.5815781354904175,
+      "learning_rate": 9.206435243478304e-06,
+      "loss": 0.1967,
+      "step": 15680
+    },
+    {
+      "epoch": 0.2066115702479339,
+      "grad_norm": 1.1220072507858276,
+      "learning_rate": 9.204129610902245e-06,
+      "loss": 0.2066,
+      "step": 15700
+    },
+    {
+      "epoch": 0.20687476970047902,
+      "grad_norm": 1.6707308292388916,
+      "learning_rate": 9.201820923358978e-06,
+      "loss": 0.2,
+      "step": 15720
+    },
+    {
+      "epoch": 0.20713796915302415,
+      "grad_norm": 1.085893988609314,
+      "learning_rate": 9.199509182526134e-06,
+      "loss": 0.2164,
+      "step": 15740
+    },
+    {
+      "epoch": 0.2074011686055693,
+      "grad_norm": 1.2889034748077393,
+      "learning_rate": 9.197194390083556e-06,
+      "loss": 0.1963,
+      "step": 15760
+    },
+    {
+      "epoch": 0.20766436805811445,
+      "grad_norm": 1.7660715579986572,
+      "learning_rate": 9.19487654771331e-06,
+      "loss": 0.1928,
+      "step": 15780
+    },
+    {
+      "epoch": 0.20792756751065958,
+      "grad_norm": 1.2228960990905762,
+      "learning_rate": 9.192555657099676e-06,
+      "loss": 0.2142,
+      "step": 15800
+    },
+    {
+      "epoch": 0.2081907669632047,
+      "grad_norm": 1.1677576303482056,
+      "learning_rate": 9.19023171992915e-06,
+      "loss": 0.2224,
+      "step": 15820
+    },
+    {
+      "epoch": 0.20845396641574984,
+      "grad_norm": 1.0813021659851074,
+      "learning_rate": 9.18790473789044e-06,
+      "loss": 0.1991,
+      "step": 15840
+    },
+    {
+      "epoch": 0.208717165868295,
+      "grad_norm": 1.6857489347457886,
+      "learning_rate": 9.185574712674468e-06,
+      "loss": 0.2236,
+      "step": 15860
+    },
+    {
+      "epoch": 0.20898036532084013,
+      "grad_norm": 1.4620684385299683,
+      "learning_rate": 9.183241645974365e-06,
+      "loss": 0.223,
+      "step": 15880
+    },
+    {
+      "epoch": 0.20924356477338527,
+      "grad_norm": 1.5597748756408691,
+      "learning_rate": 9.180905539485478e-06,
+      "loss": 0.2131,
+      "step": 15900
+    },
+    {
+      "epoch": 0.2095067642259304,
+      "grad_norm": 1.3092726469039917,
+      "learning_rate": 9.178566394905357e-06,
+      "loss": 0.2071,
+      "step": 15920
+    },
+    {
+      "epoch": 0.20976996367847556,
+      "grad_norm": 1.4240467548370361,
+      "learning_rate": 9.176224213933762e-06,
+      "loss": 0.2303,
+      "step": 15940
+    },
+    {
+      "epoch": 0.2100331631310207,
+      "grad_norm": 1.5322593450546265,
+      "learning_rate": 9.173878998272659e-06,
+      "loss": 0.1933,
+      "step": 15960
+    },
+    {
+      "epoch": 0.21029636258356582,
+      "grad_norm": 1.158003568649292,
+      "learning_rate": 9.171530749626219e-06,
+      "loss": 0.2167,
+      "step": 15980
+    },
+    {
+      "epoch": 0.21055956203611095,
+      "grad_norm": 1.2979156970977783,
+      "learning_rate": 9.169179469700816e-06,
+      "loss": 0.2058,
+      "step": 16000
+    },
+    {
+      "epoch": 0.21055956203611095,
+      "eval_loss": 0.23725290596485138,
+      "eval_runtime": 4.6637,
+      "eval_samples_per_second": 3.431,
+      "eval_steps_per_second": 1.715,
+      "step": 16000
+    },
+    {
+      "epoch": 0.2108227614886561,
+      "grad_norm": 1.6349668502807617,
+      "learning_rate": 9.16682516020503e-06,
+      "loss": 0.2115,
+      "step": 16020
+    },
+    {
+      "epoch": 0.21108596094120124,
+      "grad_norm": 1.3457437753677368,
+      "learning_rate": 9.164467822849641e-06,
+      "loss": 0.2172,
+      "step": 16040
+    },
+    {
+      "epoch": 0.21134916039374638,
+      "grad_norm": 1.2001378536224365,
+      "learning_rate": 9.162107459347627e-06,
+      "loss": 0.1779,
+      "step": 16060
+    },
+    {
+      "epoch": 0.2116123598462915,
+      "grad_norm": 2.0379481315612793,
+      "learning_rate": 9.159744071414166e-06,
+      "loss": 0.204,
+      "step": 16080
+    },
+    {
+      "epoch": 0.21187555929883667,
+      "grad_norm": 1.4113926887512207,
+      "learning_rate": 9.157377660766635e-06,
+      "loss": 0.2259,
+      "step": 16100
+    },
+    {
+      "epoch": 0.2121387587513818,
+      "grad_norm": 1.7951005697250366,
+      "learning_rate": 9.15500822912461e-06,
+      "loss": 0.2093,
+      "step": 16120
+    },
+    {
+      "epoch": 0.21240195820392693,
+      "grad_norm": 1.438250184059143,
+      "learning_rate": 9.152635778209853e-06,
+      "loss": 0.2157,
+      "step": 16140
+    },
+    {
+      "epoch": 0.21266515765647206,
+      "grad_norm": 1.519720435142517,
+      "learning_rate": 9.150260309746329e-06,
+      "loss": 0.2147,
+      "step": 16160
+    },
+    {
+      "epoch": 0.21292835710901722,
+      "grad_norm": 1.5410311222076416,
+      "learning_rate": 9.147881825460194e-06,
+      "loss": 0.2098,
+      "step": 16180
+    },
+    {
+      "epoch": 0.21319155656156236,
+      "grad_norm": 1.4750993251800537,
+      "learning_rate": 9.145500327079793e-06,
+      "loss": 0.1908,
+      "step": 16200
+    },
+    {
+      "epoch": 0.2134547560141075,
+      "grad_norm": 1.510805368423462,
+      "learning_rate": 9.143115816335663e-06,
+      "loss": 0.2107,
+      "step": 16220
+    },
+    {
+      "epoch": 0.21371795546665262,
+      "grad_norm": 1.5661991834640503,
+      "learning_rate": 9.140728294960529e-06,
+      "loss": 0.2004,
+      "step": 16240
+    },
+    {
+      "epoch": 0.21398115491919778,
+      "grad_norm": 1.452976942062378,
+      "learning_rate": 9.138337764689303e-06,
+      "loss": 0.2271,
+      "step": 16260
+    },
+    {
+      "epoch": 0.2142443543717429,
+      "grad_norm": 1.7748302221298218,
+      "learning_rate": 9.135944227259084e-06,
+      "loss": 0.1928,
+      "step": 16280
+    },
+    {
+      "epoch": 0.21450755382428804,
+      "grad_norm": 1.3658555746078491,
+      "learning_rate": 9.13354768440916e-06,
+      "loss": 0.2019,
+      "step": 16300
+    },
+    {
+      "epoch": 0.21477075327683318,
+      "grad_norm": 1.437667965888977,
+      "learning_rate": 9.131148137880998e-06,
+      "loss": 0.2193,
+      "step": 16320
+    },
+    {
+      "epoch": 0.21503395272937834,
+      "grad_norm": 1.6827822923660278,
+      "learning_rate": 9.128745589418248e-06,
+      "loss": 0.1933,
+      "step": 16340
+    },
+    {
+      "epoch": 0.21529715218192347,
+      "grad_norm": 1.232244849205017,
+      "learning_rate": 9.126340040766746e-06,
+      "loss": 0.2204,
+      "step": 16360
+    },
+    {
+      "epoch": 0.2155603516344686,
+      "grad_norm": 1.3516398668289185,
+      "learning_rate": 9.1239314936745e-06,
+      "loss": 0.1951,
+      "step": 16380
+    },
+    {
+      "epoch": 0.21582355108701373,
+      "grad_norm": 1.15445876121521,
+      "learning_rate": 9.121519949891706e-06,
+      "loss": 0.1886,
+      "step": 16400
+    },
+    {
+      "epoch": 0.21608675053955886,
+      "grad_norm": 0.9587343335151672,
+      "learning_rate": 9.11910541117073e-06,
+      "loss": 0.1983,
+      "step": 16420
+    },
+    {
+      "epoch": 0.21634994999210402,
+      "grad_norm": 1.4562147855758667,
+      "learning_rate": 9.11668787926612e-06,
+      "loss": 0.1924,
+      "step": 16440
+    },
+    {
+      "epoch": 0.21661314944464916,
+      "grad_norm": 1.4496114253997803,
+      "learning_rate": 9.114267355934595e-06,
+      "loss": 0.2086,
+      "step": 16460
+    },
+    {
+      "epoch": 0.2168763488971943,
+      "grad_norm": 1.318701148033142,
+      "learning_rate": 9.111843842935048e-06,
+      "loss": 0.2023,
+      "step": 16480
+    },
+    {
+      "epoch": 0.21713954834973942,
+      "grad_norm": 1.3021790981292725,
+      "learning_rate": 9.10941734202855e-06,
+      "loss": 0.2008,
+      "step": 16500
+    },
+    {
+      "epoch": 0.21713954834973942,
+      "eval_loss": 0.23212799429893494,
+      "eval_runtime": 3.7213,
+      "eval_samples_per_second": 4.3,
+      "eval_steps_per_second": 2.15,
+      "step": 16500
+    },
+    {
+      "epoch": 0.21740274780228458,
+      "grad_norm": 1.2814044952392578,
+      "learning_rate": 9.106987854978334e-06,
+      "loss": 0.2207,
+      "step": 16520
+    },
+    {
+      "epoch": 0.2176659472548297,
+      "grad_norm": 1.6028743982315063,
+      "learning_rate": 9.104555383549811e-06,
+      "loss": 0.1982,
+      "step": 16540
+    },
+    {
+      "epoch": 0.21792914670737484,
+      "grad_norm": 1.4128284454345703,
+      "learning_rate": 9.102119929510555e-06,
+      "loss": 0.1851,
+      "step": 16560
+    },
+    {
+      "epoch": 0.21819234615991998,
+      "grad_norm": 1.200227975845337,
+      "learning_rate": 9.099681494630312e-06,
+      "loss": 0.2155,
+      "step": 16580
+    },
+    {
+      "epoch": 0.21845554561246514,
+      "grad_norm": 1.3398634195327759,
+      "learning_rate": 9.097240080680992e-06,
+      "loss": 0.2016,
+      "step": 16600
+    },
+    {
+      "epoch": 0.21871874506501027,
+      "grad_norm": 1.2564581632614136,
+      "learning_rate": 9.094795689436668e-06,
+      "loss": 0.1984,
+      "step": 16620
+    },
+    {
+      "epoch": 0.2189819445175554,
+      "grad_norm": 2.386012077331543,
+      "learning_rate": 9.092348322673578e-06,
+      "loss": 0.2037,
+      "step": 16640
+    },
+    {
+      "epoch": 0.21924514397010053,
+      "grad_norm": 1.1463801860809326,
+      "learning_rate": 9.089897982170126e-06,
+      "loss": 0.2096,
+      "step": 16660
+    },
+    {
+      "epoch": 0.2195083434226457,
+      "grad_norm": 1.7645434141159058,
+      "learning_rate": 9.087444669706868e-06,
+      "loss": 0.2056,
+      "step": 16680
+    },
+    {
+      "epoch": 0.21977154287519082,
+      "grad_norm": 1.5234774351119995,
+      "learning_rate": 9.084988387066527e-06,
+      "loss": 0.2078,
+      "step": 16700
+    },
+    {
+      "epoch": 0.22003474232773595,
+      "grad_norm": 1.1727930307388306,
+      "learning_rate": 9.082529136033982e-06,
+      "loss": 0.2057,
+      "step": 16720
+    },
+    {
+      "epoch": 0.2202979417802811,
+      "grad_norm": 1.3146255016326904,
+      "learning_rate": 9.080066918396268e-06,
+      "loss": 0.1941,
+      "step": 16740
+    },
+    {
+      "epoch": 0.22056114123282625,
+      "grad_norm": 1.1046918630599976,
+      "learning_rate": 9.077601735942578e-06,
+      "loss": 0.2155,
+      "step": 16760
+    },
+    {
+      "epoch": 0.22082434068537138,
+      "grad_norm": 1.3277379274368286,
+      "learning_rate": 9.07513359046426e-06,
+      "loss": 0.2029,
+      "step": 16780
+    },
+    {
+      "epoch": 0.2210875401379165,
+      "grad_norm": 1.3440901041030884,
+      "learning_rate": 9.072662483754809e-06,
+      "loss": 0.1966,
+      "step": 16800
+    },
+    {
+      "epoch": 0.22135073959046164,
+      "grad_norm": 1.4013023376464844,
+      "learning_rate": 9.07018841760988e-06,
+      "loss": 0.2101,
+      "step": 16820
+    },
+    {
+      "epoch": 0.2216139390430068,
+      "grad_norm": 1.3899908065795898,
+      "learning_rate": 9.067711393827272e-06,
+      "loss": 0.2223,
+      "step": 16840
+    },
+    {
+      "epoch": 0.22187713849555193,
+      "grad_norm": 1.5671601295471191,
+      "learning_rate": 9.065231414206937e-06,
+      "loss": 0.2047,
+      "step": 16860
+    },
+    {
+      "epoch": 0.22214033794809707,
+      "grad_norm": 1.5202046632766724,
+      "learning_rate": 9.062748480550973e-06,
+      "loss": 0.2047,
+      "step": 16880
+    },
+    {
+      "epoch": 0.2224035374006422,
+      "grad_norm": 1.407057762145996,
+      "learning_rate": 9.060262594663627e-06,
+      "loss": 0.21,
+      "step": 16900
+    },
+    {
+      "epoch": 0.22266673685318736,
+      "grad_norm": 1.6400309801101685,
+      "learning_rate": 9.057773758351289e-06,
+      "loss": 0.2106,
+      "step": 16920
+    },
+    {
+      "epoch": 0.2229299363057325,
+      "grad_norm": 1.5609228610992432,
+      "learning_rate": 9.05528197342249e-06,
+      "loss": 0.2041,
+      "step": 16940
+    },
+    {
+      "epoch": 0.22319313575827762,
+      "grad_norm": 1.3106389045715332,
+      "learning_rate": 9.05278724168791e-06,
+      "loss": 0.2143,
+      "step": 16960
+    },
+    {
+      "epoch": 0.22345633521082275,
+      "grad_norm": 1.0952203273773193,
+      "learning_rate": 9.05028956496037e-06,
+      "loss": 0.2205,
+      "step": 16980
+    },
+    {
+      "epoch": 0.2237195346633679,
+      "grad_norm": 1.6061345338821411,
+      "learning_rate": 9.047788945054823e-06,
+      "loss": 0.1803,
+      "step": 17000
+    },
+    {
+      "epoch": 0.2237195346633679,
+      "eval_loss": 0.23000529408454895,
+      "eval_runtime": 4.0475,
+      "eval_samples_per_second": 3.953,
+      "eval_steps_per_second": 1.977,
+      "step": 17000
+    },
+    {
+      "epoch": 0.22398273411591305,
+      "grad_norm": 1.611167311668396,
+      "learning_rate": 9.04528538378837e-06,
+      "loss": 0.2127,
+      "step": 17020
+    },
+    {
+      "epoch": 0.22424593356845818,
+      "grad_norm": 1.4692379236221313,
+      "learning_rate": 9.042778882980244e-06,
+      "loss": 0.2066,
+      "step": 17040
+    },
+    {
+      "epoch": 0.2245091330210033,
+      "grad_norm": 1.5057069063186646,
+      "learning_rate": 9.040269444451817e-06,
+      "loss": 0.2021,
+      "step": 17060
+    },
+    {
+      "epoch": 0.22477233247354844,
+      "grad_norm": 1.2186074256896973,
+      "learning_rate": 9.037757070026591e-06,
+      "loss": 0.2062,
+      "step": 17080
+    },
+    {
+      "epoch": 0.2250355319260936,
+      "grad_norm": 1.3339877128601074,
+      "learning_rate": 9.035241761530208e-06,
+      "loss": 0.2044,
+      "step": 17100
+    },
+    {
+      "epoch": 0.22529873137863873,
+      "grad_norm": 1.4815906286239624,
+      "learning_rate": 9.032723520790437e-06,
+      "loss": 0.1826,
+      "step": 17120
+    },
+    {
+      "epoch": 0.22556193083118387,
+      "grad_norm": 1.3250195980072021,
+      "learning_rate": 9.03020234963718e-06,
+      "loss": 0.2255,
+      "step": 17140
+    },
+    {
+      "epoch": 0.225825130283729,
+      "grad_norm": 1.215611457824707,
+      "learning_rate": 9.027678249902468e-06,
+      "loss": 0.2153,
+      "step": 17160
+    },
+    {
+      "epoch": 0.22608832973627416,
+      "grad_norm": 1.3328697681427002,
+      "learning_rate": 9.025151223420458e-06,
+      "loss": 0.1954,
+      "step": 17180
+    },
+    {
+      "epoch": 0.2263515291888193,
+      "grad_norm": 1.281994342803955,
+      "learning_rate": 9.022621272027441e-06,
+      "loss": 0.213,
+      "step": 17200
+    },
+    {
+      "epoch": 0.22661472864136442,
+      "grad_norm": 1.5722020864486694,
+      "learning_rate": 9.020088397561821e-06,
+      "loss": 0.2042,
+      "step": 17220
+    },
+    {
+      "epoch": 0.22687792809390955,
+      "grad_norm": 1.434576153755188,
+      "learning_rate": 9.01755260186414e-06,
+      "loss": 0.1765,
+      "step": 17240
+    },
+    {
+      "epoch": 0.2271411275464547,
+      "grad_norm": 1.6449817419052124,
+      "learning_rate": 9.01501388677705e-06,
+      "loss": 0.2035,
+      "step": 17260
+    },
+    {
+      "epoch": 0.22740432699899985,
+      "grad_norm": 1.2270034551620483,
+      "learning_rate": 9.012472254145333e-06,
+      "loss": 0.2032,
+      "step": 17280
+    },
+    {
+      "epoch": 0.22766752645154498,
+      "grad_norm": 1.7301334142684937,
+      "learning_rate": 9.00992770581589e-06,
+      "loss": 0.1851,
+      "step": 17300
+    },
+    {
+      "epoch": 0.2279307259040901,
+      "grad_norm": 1.3540586233139038,
+      "learning_rate": 9.007380243637733e-06,
+      "loss": 0.196,
+      "step": 17320
+    },
+    {
+      "epoch": 0.22819392535663527,
+      "grad_norm": 1.3233907222747803,
+      "learning_rate": 9.004829869462005e-06,
+      "loss": 0.1927,
+      "step": 17340
+    },
+    {
+      "epoch": 0.2284571248091804,
+      "grad_norm": 1.4068877696990967,
+      "learning_rate": 9.002276585141954e-06,
+      "loss": 0.2055,
+      "step": 17360
+    },
+    {
+      "epoch": 0.22872032426172553,
+      "grad_norm": 1.269504189491272,
+      "learning_rate": 8.999720392532946e-06,
+      "loss": 0.2272,
+      "step": 17380
+    },
+    {
+      "epoch": 0.22898352371427066,
+      "grad_norm": 1.5798684358596802,
+      "learning_rate": 8.997161293492462e-06,
+      "loss": 0.2073,
+      "step": 17400
+    },
+    {
+      "epoch": 0.22924672316681582,
+      "grad_norm": 1.3535726070404053,
+      "learning_rate": 8.994599289880091e-06,
+      "loss": 0.2038,
+      "step": 17420
+    },
+    {
+      "epoch": 0.22950992261936096,
+      "grad_norm": 1.289067029953003,
+      "learning_rate": 8.992034383557538e-06,
+      "loss": 0.2027,
+      "step": 17440
+    },
+    {
+      "epoch": 0.2297731220719061,
+      "grad_norm": 1.3632245063781738,
+      "learning_rate": 8.989466576388612e-06,
+      "loss": 0.2088,
+      "step": 17460
+    },
+    {
+      "epoch": 0.23003632152445122,
+      "grad_norm": 1.3355331420898438,
+      "learning_rate": 8.986895870239234e-06,
+      "loss": 0.1998,
+      "step": 17480
+    },
+    {
+      "epoch": 0.23029952097699638,
+      "grad_norm": 1.543005108833313,
+      "learning_rate": 8.98432226697743e-06,
+      "loss": 0.1901,
+      "step": 17500
+    },
+    {
+      "epoch": 0.23029952097699638,
+      "eval_loss": 0.22878733277320862,
+      "eval_runtime": 3.9749,
+      "eval_samples_per_second": 4.025,
+      "eval_steps_per_second": 2.013,
+      "step": 17500
+    },
+    {
+      "epoch": 0.2305627204295415,
+      "grad_norm": 1.4697489738464355,
+      "learning_rate": 8.981745768473329e-06,
+      "loss": 0.1995,
+      "step": 17520
+    },
+    {
+      "epoch": 0.23082591988208664,
+      "grad_norm": 1.7296479940414429,
+      "learning_rate": 8.979166376599167e-06,
+      "loss": 0.1925,
+      "step": 17540
+    },
+    {
+      "epoch": 0.23108911933463178,
+      "grad_norm": 1.661498785018921,
+      "learning_rate": 8.976584093229282e-06,
+      "loss": 0.1982,
+      "step": 17560
+    },
+    {
+      "epoch": 0.23135231878717694,
+      "grad_norm": 1.2969017028808594,
+      "learning_rate": 8.973998920240113e-06,
+      "loss": 0.1861,
+      "step": 17580
+    },
+    {
+      "epoch": 0.23161551823972207,
+      "grad_norm": 1.586206316947937,
+      "learning_rate": 8.971410859510197e-06,
+      "loss": 0.1893,
+      "step": 17600
+    },
+    {
+      "epoch": 0.2318787176922672,
+      "grad_norm": 1.3976099491119385,
+      "learning_rate": 8.96881991292017e-06,
+      "loss": 0.2006,
+      "step": 17620
+    },
+    {
+      "epoch": 0.23214191714481233,
+      "grad_norm": 1.2380386590957642,
+      "learning_rate": 8.96622608235277e-06,
+      "loss": 0.1996,
+      "step": 17640
+    },
+    {
+      "epoch": 0.23240511659735746,
+      "grad_norm": 1.399925947189331,
+      "learning_rate": 8.963629369692821e-06,
+      "loss": 0.2094,
+      "step": 17660
+    },
+    {
+      "epoch": 0.23266831604990262,
+      "grad_norm": 1.2759567499160767,
+      "learning_rate": 8.961029776827252e-06,
+      "loss": 0.196,
+      "step": 17680
+    },
+    {
+      "epoch": 0.23293151550244776,
+      "grad_norm": 1.280727505683899,
+      "learning_rate": 8.958427305645076e-06,
+      "loss": 0.1901,
+      "step": 17700
+    },
+    {
+      "epoch": 0.2331947149549929,
+      "grad_norm": 1.143575668334961,
+      "learning_rate": 8.955821958037402e-06,
+      "loss": 0.1902,
+      "step": 17720
+    },
+    {
+      "epoch": 0.23345791440753802,
+      "grad_norm": 1.7547069787979126,
+      "learning_rate": 8.953213735897428e-06,
+      "loss": 0.1876,
+      "step": 17740
+    },
+    {
+      "epoch": 0.23372111386008318,
+      "grad_norm": 1.25863516330719,
+      "learning_rate": 8.950602641120443e-06,
+      "loss": 0.1965,
+      "step": 17760
+    },
+    {
+      "epoch": 0.2339843133126283,
+      "grad_norm": 1.194683313369751,
+      "learning_rate": 8.94798867560382e-06,
+      "loss": 0.2151,
+      "step": 17780
+    },
+    {
+      "epoch": 0.23424751276517344,
+      "grad_norm": 1.5396754741668701,
+      "learning_rate": 8.94537184124702e-06,
+      "loss": 0.1849,
+      "step": 17800
+    },
+    {
+      "epoch": 0.23451071221771858,
+      "grad_norm": 1.2771893739700317,
+      "learning_rate": 8.94275213995159e-06,
+      "loss": 0.1965,
+      "step": 17820
+    },
+    {
+      "epoch": 0.23477391167026374,
+      "grad_norm": 1.029513955116272,
+      "learning_rate": 8.940129573621156e-06,
+      "loss": 0.2181,
+      "step": 17840
+    },
+    {
+      "epoch": 0.23503711112280887,
+      "grad_norm": 1.2193193435668945,
+      "learning_rate": 8.937504144161431e-06,
+      "loss": 0.2141,
+      "step": 17860
+    },
+    {
+      "epoch": 0.235300310575354,
+      "grad_norm": 1.677281379699707,
+      "learning_rate": 8.934875853480204e-06,
+      "loss": 0.1836,
+      "step": 17880
+    },
+    {
+      "epoch": 0.23556351002789913,
+      "grad_norm": 1.6836661100387573,
+      "learning_rate": 8.932244703487347e-06,
+      "loss": 0.1896,
+      "step": 17900
+    },
+    {
+      "epoch": 0.2358267094804443,
+      "grad_norm": 1.2722394466400146,
+      "learning_rate": 8.929610696094806e-06,
+      "loss": 0.209,
+      "step": 17920
+    },
+    {
+      "epoch": 0.23608990893298942,
+      "grad_norm": 1.5300798416137695,
+      "learning_rate": 8.926973833216607e-06,
+      "loss": 0.1945,
+      "step": 17940
+    },
+    {
+      "epoch": 0.23635310838553455,
+      "grad_norm": 1.5486698150634766,
+      "learning_rate": 8.924334116768848e-06,
+      "loss": 0.2035,
+      "step": 17960
+    },
+    {
+      "epoch": 0.2366163078380797,
+      "grad_norm": 1.3547779321670532,
+      "learning_rate": 8.921691548669703e-06,
+      "loss": 0.2077,
+      "step": 17980
+    },
+    {
+      "epoch": 0.23687950729062485,
+      "grad_norm": 1.4039422273635864,
+      "learning_rate": 8.919046130839415e-06,
+      "loss": 0.2121,
+      "step": 18000
+    },
+    {
+      "epoch": 0.23687950729062485,
+      "eval_loss": 0.23066562414169312,
+      "eval_runtime": 4.0518,
+      "eval_samples_per_second": 3.949,
+      "eval_steps_per_second": 1.974,
+      "step": 18000
+    },
+    {
+      "epoch": 0.23714270674316998,
+      "grad_norm": 1.9173293113708496,
+      "learning_rate": 8.916397865200303e-06,
+      "loss": 0.1888,
+      "step": 18020
+    },
+    {
+      "epoch": 0.2374059061957151,
+      "grad_norm": 1.4101506471633911,
+      "learning_rate": 8.91374675367675e-06,
+      "loss": 0.212,
+      "step": 18040
+    },
+    {
+      "epoch": 0.23766910564826024,
+      "grad_norm": 1.3767117261886597,
+      "learning_rate": 8.91109279819521e-06,
+      "loss": 0.2024,
+      "step": 18060
+    },
+    {
+      "epoch": 0.2379323051008054,
+      "grad_norm": 1.4015063047409058,
+      "learning_rate": 8.9084360006842e-06,
+      "loss": 0.1958,
+      "step": 18080
+    },
+    {
+      "epoch": 0.23819550455335053,
+      "grad_norm": 1.6327142715454102,
+      "learning_rate": 8.90577636307431e-06,
+      "loss": 0.1892,
+      "step": 18100
+    },
+    {
+      "epoch": 0.23845870400589567,
+      "grad_norm": 1.533033847808838,
+      "learning_rate": 8.903113887298183e-06,
+      "loss": 0.2057,
+      "step": 18120
+    },
+    {
+      "epoch": 0.2387219034584408,
+      "grad_norm": 1.3615649938583374,
+      "learning_rate": 8.900448575290534e-06,
+      "loss": 0.2036,
+      "step": 18140
+    },
+    {
+      "epoch": 0.23898510291098596,
+      "grad_norm": 1.8697692155838013,
+      "learning_rate": 8.897780428988137e-06,
+      "loss": 0.1963,
+      "step": 18160
+    },
+    {
+      "epoch": 0.2392483023635311,
+      "grad_norm": 1.212471604347229,
+      "learning_rate": 8.895109450329817e-06,
+      "loss": 0.2097,
+      "step": 18180
+    },
+    {
+      "epoch": 0.23951150181607622,
+      "grad_norm": 1.563098669052124,
+      "learning_rate": 8.892435641256468e-06,
+      "loss": 0.1871,
+      "step": 18200
+    },
+    {
+      "epoch": 0.23977470126862135,
+      "grad_norm": 1.4251550436019897,
+      "learning_rate": 8.889759003711036e-06,
+      "loss": 0.2059,
+      "step": 18220
+    },
+    {
+      "epoch": 0.2400379007211665,
+      "grad_norm": 1.5386689901351929,
+      "learning_rate": 8.887079539638525e-06,
+      "loss": 0.2218,
+      "step": 18240
+    },
+    {
+      "epoch": 0.24030110017371165,
+      "grad_norm": 1.5487643480300903,
+      "learning_rate": 8.884397250985987e-06,
+      "loss": 0.1959,
+      "step": 18260
+    },
+    {
+      "epoch": 0.24056429962625678,
+      "grad_norm": 1.4674370288848877,
+      "learning_rate": 8.881712139702533e-06,
+      "loss": 0.1878,
+      "step": 18280
+    },
+    {
+      "epoch": 0.2408274990788019,
+      "grad_norm": 1.5473144054412842,
+      "learning_rate": 8.87902420773932e-06,
+      "loss": 0.2235,
+      "step": 18300
+    },
+    {
+      "epoch": 0.24109069853134704,
+      "grad_norm": 1.8335022926330566,
+      "learning_rate": 8.876333457049558e-06,
+      "loss": 0.2061,
+      "step": 18320
+    },
+    {
+      "epoch": 0.2413538979838922,
+      "grad_norm": 1.3688385486602783,
+      "learning_rate": 8.873639889588505e-06,
+      "loss": 0.1839,
+      "step": 18340
+    },
+    {
+      "epoch": 0.24161709743643733,
+      "grad_norm": 1.3568267822265625,
+      "learning_rate": 8.870943507313466e-06,
+      "loss": 0.1986,
+      "step": 18360
+    },
+    {
+      "epoch": 0.24188029688898247,
+      "grad_norm": 1.234585165977478,
+      "learning_rate": 8.868244312183787e-06,
+      "loss": 0.1933,
+      "step": 18380
+    },
+    {
+      "epoch": 0.2421434963415276,
+      "grad_norm": 1.161259651184082,
+      "learning_rate": 8.865542306160865e-06,
+      "loss": 0.2045,
+      "step": 18400
+    },
+    {
+      "epoch": 0.24240669579407276,
+      "grad_norm": 1.7092849016189575,
+      "learning_rate": 8.862837491208137e-06,
+      "loss": 0.1996,
+      "step": 18420
+    },
+    {
+      "epoch": 0.2426698952466179,
+      "grad_norm": 1.4544683694839478,
+      "learning_rate": 8.860129869291077e-06,
+      "loss": 0.1851,
+      "step": 18440
+    },
+    {
+      "epoch": 0.24293309469916302,
+      "grad_norm": 1.6256581544876099,
+      "learning_rate": 8.857419442377202e-06,
+      "loss": 0.2147,
+      "step": 18460
+    },
+    {
+      "epoch": 0.24319629415170815,
+      "grad_norm": 1.8777192831039429,
+      "learning_rate": 8.854706212436068e-06,
+      "loss": 0.1891,
+      "step": 18480
+    },
+    {
+      "epoch": 0.2434594936042533,
+      "grad_norm": 2.3471803665161133,
+      "learning_rate": 8.851990181439269e-06,
+      "loss": 0.193,
+      "step": 18500
+    },
+    {
+      "epoch": 0.2434594936042533,
+      "eval_loss": 0.2336764633655548,
+      "eval_runtime": 3.3432,
+      "eval_samples_per_second": 4.786,
+      "eval_steps_per_second": 2.393,
+      "step": 18500
+    },
+    {
+      "epoch": 0.24372269305679845,
+      "grad_norm": 1.0334043502807617,
+      "learning_rate": 8.849271351360431e-06,
+      "loss": 0.2079,
+      "step": 18520
+    },
+    {
+      "epoch": 0.24398589250934358,
+      "grad_norm": 1.2378817796707153,
+      "learning_rate": 8.846549724175214e-06,
+      "loss": 0.2065,
+      "step": 18540
+    },
+    {
+      "epoch": 0.2442490919618887,
+      "grad_norm": 1.410272479057312,
+      "learning_rate": 8.843825301861312e-06,
+      "loss": 0.2006,
+      "step": 18560
+    },
+    {
+      "epoch": 0.24451229141443387,
+      "grad_norm": 1.372580647468567,
+      "learning_rate": 8.841098086398454e-06,
+      "loss": 0.1852,
+      "step": 18580
+    },
+    {
+      "epoch": 0.244775490866979,
+      "grad_norm": 1.2856806516647339,
+      "learning_rate": 8.838368079768392e-06,
+      "loss": 0.2062,
+      "step": 18600
+    },
+    {
+      "epoch": 0.24503869031952413,
+      "grad_norm": 1.4348926544189453,
+      "learning_rate": 8.835635283954907e-06,
+      "loss": 0.1863,
+      "step": 18620
+    },
+    {
+      "epoch": 0.24530188977206926,
+      "grad_norm": 1.8706684112548828,
+      "learning_rate": 8.832899700943811e-06,
+      "loss": 0.2066,
+      "step": 18640
+    },
+    {
+      "epoch": 0.24556508922461442,
+      "grad_norm": 1.4998947381973267,
+      "learning_rate": 8.83016133272294e-06,
+      "loss": 0.2077,
+      "step": 18660
+    },
+    {
+      "epoch": 0.24582828867715956,
+      "grad_norm": 1.238265037536621,
+      "learning_rate": 8.827420181282152e-06,
+      "loss": 0.1974,
+      "step": 18680
+    },
+    {
+      "epoch": 0.2460914881297047,
+      "grad_norm": 1.2517786026000977,
+      "learning_rate": 8.824676248613326e-06,
+      "loss": 0.1987,
+      "step": 18700
+    },
+    {
+      "epoch": 0.24635468758224982,
+      "grad_norm": 1.2829886674880981,
+      "learning_rate": 8.821929536710369e-06,
+      "loss": 0.1813,
+      "step": 18720
+    },
+    {
+      "epoch": 0.24661788703479498,
+      "grad_norm": 1.5087491273880005,
+      "learning_rate": 8.8191800475692e-06,
+      "loss": 0.203,
+      "step": 18740
+    },
+    {
+      "epoch": 0.2468810864873401,
+      "grad_norm": 0.9559106826782227,
+      "learning_rate": 8.81642778318776e-06,
+      "loss": 0.192,
+      "step": 18760
+    },
+    {
+      "epoch": 0.24714428593988524,
+      "grad_norm": 1.549684762954712,
+      "learning_rate": 8.813672745566005e-06,
+      "loss": 0.1871,
+      "step": 18780
+    },
+    {
+      "epoch": 0.24740748539243038,
+      "grad_norm": 1.4092354774475098,
+      "learning_rate": 8.810914936705907e-06,
+      "loss": 0.1897,
+      "step": 18800
+    },
+    {
+      "epoch": 0.24767068484497554,
+      "grad_norm": 0.9358004927635193,
+      "learning_rate": 8.808154358611453e-06,
+      "loss": 0.1771,
+      "step": 18820
+    },
+    {
+      "epoch": 0.24793388429752067,
+      "grad_norm": 1.2603906393051147,
+      "learning_rate": 8.805391013288642e-06,
+      "loss": 0.1993,
+      "step": 18840
+    },
+    {
+      "epoch": 0.2481970837500658,
+      "grad_norm": 1.2440749406814575,
+      "learning_rate": 8.802624902745479e-06,
+      "loss": 0.2054,
+      "step": 18860
+    },
+    {
+      "epoch": 0.24846028320261093,
+      "grad_norm": 1.4302537441253662,
+      "learning_rate": 8.799856028991984e-06,
+      "loss": 0.1915,
+      "step": 18880
+    },
+    {
+      "epoch": 0.24872348265515606,
+      "grad_norm": 1.7404459714889526,
+      "learning_rate": 8.797084394040185e-06,
+      "loss": 0.1989,
+      "step": 18900
+    },
+    {
+      "epoch": 0.24898668210770122,
+      "grad_norm": 1.2364033460617065,
+      "learning_rate": 8.794309999904111e-06,
+      "loss": 0.1963,
+      "step": 18920
+    },
+    {
+      "epoch": 0.24924988156024636,
+      "grad_norm": 1.8852834701538086,
+      "learning_rate": 8.791532848599804e-06,
+      "loss": 0.1887,
+      "step": 18940
+    },
+    {
+      "epoch": 0.2495130810127915,
+      "grad_norm": 1.3845535516738892,
+      "learning_rate": 8.7887529421453e-06,
+      "loss": 0.2055,
+      "step": 18960
+    },
+    {
+      "epoch": 0.24977628046533662,
+      "grad_norm": 1.3393343687057495,
+      "learning_rate": 8.785970282560647e-06,
+      "loss": 0.1918,
+      "step": 18980
+    },
+    {
+      "epoch": 0.2500394799178818,
+      "grad_norm": 1.8313708305358887,
+      "learning_rate": 8.783184871867884e-06,
+      "loss": 0.2142,
+      "step": 19000
+    },
+    {
+      "epoch": 0.2500394799178818,
+      "eval_loss": 0.22771388292312622,
+      "eval_runtime": 4.2145,
+      "eval_samples_per_second": 3.796,
+      "eval_steps_per_second": 1.898,
+      "step": 19000
+    },
+    {
+      "epoch": 0.2503026793704269,
+      "grad_norm": 1.792867660522461,
+      "learning_rate": 8.780396712091057e-06,
+      "loss": 0.1862,
+      "step": 19020
+    },
+    {
+      "epoch": 0.25056587882297204,
+      "grad_norm": 1.111887812614441,
+      "learning_rate": 8.777605805256206e-06,
+      "loss": 0.1995,
+      "step": 19040
+    },
+    {
+      "epoch": 0.2508290782755172,
+      "grad_norm": 1.4841375350952148,
+      "learning_rate": 8.774812153391367e-06,
+      "loss": 0.1969,
+      "step": 19060
+    },
+    {
+      "epoch": 0.2510922777280623,
+      "grad_norm": 1.5620406866073608,
+      "learning_rate": 8.772015758526572e-06,
+      "loss": 0.2212,
+      "step": 19080
+    },
+    {
+      "epoch": 0.25135547718060747,
+      "grad_norm": 1.344254970550537,
+      "learning_rate": 8.769216622693843e-06,
+      "loss": 0.215,
+      "step": 19100
+    },
+    {
+      "epoch": 0.2516186766331526,
+      "grad_norm": 1.1537009477615356,
+      "learning_rate": 8.7664147479272e-06,
+      "loss": 0.1989,
+      "step": 19120
+    },
+    {
+      "epoch": 0.25188187608569773,
+      "grad_norm": 1.526115894317627,
+      "learning_rate": 8.763610136262649e-06,
+      "loss": 0.219,
+      "step": 19140
+    },
+    {
+      "epoch": 0.2521450755382429,
+      "grad_norm": 1.8861950635910034,
+      "learning_rate": 8.760802789738183e-06,
+      "loss": 0.2067,
+      "step": 19160
+    },
+    {
+      "epoch": 0.252408274990788,
+      "grad_norm": 1.411926031112671,
+      "learning_rate": 8.757992710393786e-06,
+      "loss": 0.2159,
+      "step": 19180
+    },
+    {
+      "epoch": 0.25267147444333315,
+      "grad_norm": 1.6802513599395752,
+      "learning_rate": 8.755179900271429e-06,
+      "loss": 0.199,
+      "step": 19200
+    },
+    {
+      "epoch": 0.2529346738958783,
+      "grad_norm": 1.28301203250885,
+      "learning_rate": 8.75236436141506e-06,
+      "loss": 0.191,
+      "step": 19220
+    },
+    {
+      "epoch": 0.2531978733484234,
+      "grad_norm": 1.2079546451568604,
+      "learning_rate": 8.749546095870617e-06,
+      "loss": 0.201,
+      "step": 19240
+    },
+    {
+      "epoch": 0.2534610728009686,
+      "grad_norm": 1.2686374187469482,
+      "learning_rate": 8.746725105686017e-06,
+      "loss": 0.1976,
+      "step": 19260
+    },
+    {
+      "epoch": 0.25372427225351374,
+      "grad_norm": 2.0446290969848633,
+      "learning_rate": 8.743901392911158e-06,
+      "loss": 0.1848,
+      "step": 19280
+    },
+    {
+      "epoch": 0.25398747170605884,
+      "grad_norm": 1.037169337272644,
+      "learning_rate": 8.741074959597914e-06,
+      "loss": 0.2115,
+      "step": 19300
+    },
+    {
+      "epoch": 0.254250671158604,
+      "grad_norm": 1.7340439558029175,
+      "learning_rate": 8.738245807800138e-06,
+      "loss": 0.1964,
+      "step": 19320
+    },
+    {
+      "epoch": 0.2545138706111491,
+      "grad_norm": 1.6496762037277222,
+      "learning_rate": 8.735413939573657e-06,
+      "loss": 0.1818,
+      "step": 19340
+    },
+    {
+      "epoch": 0.25477707006369427,
+      "grad_norm": 1.4309301376342773,
+      "learning_rate": 8.732579356976273e-06,
+      "loss": 0.214,
+      "step": 19360
+    },
+    {
+      "epoch": 0.2550402695162394,
+      "grad_norm": 1.4722577333450317,
+      "learning_rate": 8.72974206206776e-06,
+      "loss": 0.1995,
+      "step": 19380
+    },
+    {
+      "epoch": 0.25530346896878453,
+      "grad_norm": 1.5048785209655762,
+      "learning_rate": 8.726902056909863e-06,
+      "loss": 0.2191,
+      "step": 19400
+    },
+    {
+      "epoch": 0.2555666684213297,
+      "grad_norm": 1.6017286777496338,
+      "learning_rate": 8.724059343566295e-06,
+      "loss": 0.1942,
+      "step": 19420
+    },
+    {
+      "epoch": 0.25582986787387485,
+      "grad_norm": 1.6503804922103882,
+      "learning_rate": 8.721213924102743e-06,
+      "loss": 0.2061,
+      "step": 19440
+    },
+    {
+      "epoch": 0.25609306732641995,
+      "grad_norm": 1.4935156106948853,
+      "learning_rate": 8.71836580058685e-06,
+      "loss": 0.1925,
+      "step": 19460
+    },
+    {
+      "epoch": 0.2563562667789651,
+      "grad_norm": 1.4751176834106445,
+      "learning_rate": 8.71551497508823e-06,
+      "loss": 0.2108,
+      "step": 19480
+    },
+    {
+      "epoch": 0.2566194662315102,
+      "grad_norm": 1.4118298292160034,
+      "learning_rate": 8.712661449678466e-06,
+      "loss": 0.2146,
+      "step": 19500
+    },
+    {
+      "epoch": 0.2566194662315102,
+      "eval_loss": 0.22961989045143127,
+      "eval_runtime": 4.5963,
+      "eval_samples_per_second": 3.481,
+      "eval_steps_per_second": 1.741,
+      "step": 19500
+    },
+    {
+      "epoch": 0.2568826656840554,
+      "grad_norm": 1.3975489139556885,
+      "learning_rate": 8.70980522643109e-06,
+      "loss": 0.2113,
+      "step": 19520
+    },
+    {
+      "epoch": 0.25714586513660054,
+      "grad_norm": 1.3045250177383423,
+      "learning_rate": 8.706946307421605e-06,
+      "loss": 0.213,
+      "step": 19540
+    },
+    {
+      "epoch": 0.25740906458914564,
+      "grad_norm": 1.4240150451660156,
+      "learning_rate": 8.704084694727468e-06,
+      "loss": 0.2,
+      "step": 19560
+    },
+    {
+      "epoch": 0.2576722640416908,
+      "grad_norm": 1.5030370950698853,
+      "learning_rate": 8.701220390428096e-06,
+      "loss": 0.1967,
+      "step": 19580
+    },
+    {
+      "epoch": 0.2579354634942359,
+      "grad_norm": 1.4377492666244507,
+      "learning_rate": 8.698353396604859e-06,
+      "loss": 0.2074,
+      "step": 19600
+    },
+    {
+      "epoch": 0.25819866294678107,
+      "grad_norm": 1.017295479774475,
+      "learning_rate": 8.695483715341084e-06,
+      "loss": 0.1895,
+      "step": 19620
+    },
+    {
+      "epoch": 0.2584618623993262,
+      "grad_norm": 1.3737146854400635,
+      "learning_rate": 8.69261134872205e-06,
+      "loss": 0.203,
+      "step": 19640
+    },
+    {
+      "epoch": 0.25872506185187133,
+      "grad_norm": 1.3209843635559082,
+      "learning_rate": 8.689736298834985e-06,
+      "loss": 0.1912,
+      "step": 19660
+    },
+    {
+      "epoch": 0.2589882613044165,
+      "grad_norm": 1.6510685682296753,
+      "learning_rate": 8.686858567769073e-06,
+      "loss": 0.2027,
+      "step": 19680
+    },
+    {
+      "epoch": 0.25925146075696165,
+      "grad_norm": 1.6452299356460571,
+      "learning_rate": 8.683978157615435e-06,
+      "loss": 0.1901,
+      "step": 19700
+    },
+    {
+      "epoch": 0.25951466020950675,
+      "grad_norm": 1.3856523036956787,
+      "learning_rate": 8.681095070467155e-06,
+      "loss": 0.1899,
+      "step": 19720
+    },
+    {
+      "epoch": 0.2597778596620519,
+      "grad_norm": 1.4651427268981934,
+      "learning_rate": 8.678209308419247e-06,
+      "loss": 0.2031,
+      "step": 19740
+    },
+    {
+      "epoch": 0.260041059114597,
+      "grad_norm": 1.442284345626831,
+      "learning_rate": 8.675320873568676e-06,
+      "loss": 0.2064,
+      "step": 19760
+    },
+    {
+      "epoch": 0.2603042585671422,
+      "grad_norm": 1.4081627130508423,
+      "learning_rate": 8.672429768014351e-06,
+      "loss": 0.1832,
+      "step": 19780
+    },
+    {
+      "epoch": 0.26056745801968734,
+      "grad_norm": 1.509494662284851,
+      "learning_rate": 8.669535993857116e-06,
+      "loss": 0.2062,
+      "step": 19800
+    },
+    {
+      "epoch": 0.26083065747223244,
+      "grad_norm": 1.3971705436706543,
+      "learning_rate": 8.666639553199758e-06,
+      "loss": 0.2272,
+      "step": 19820
+    },
+    {
+      "epoch": 0.2610938569247776,
+      "grad_norm": 0.9510790109634399,
+      "learning_rate": 8.663740448147e-06,
+      "loss": 0.2056,
+      "step": 19840
+    },
+    {
+      "epoch": 0.26135705637732276,
+      "grad_norm": 1.1784926652908325,
+      "learning_rate": 8.660838680805504e-06,
+      "loss": 0.1884,
+      "step": 19860
+    },
+    {
+      "epoch": 0.26162025582986786,
+      "grad_norm": 1.4749836921691895,
+      "learning_rate": 8.657934253283861e-06,
+      "loss": 0.2061,
+      "step": 19880
+    },
+    {
+      "epoch": 0.261883455282413,
+      "grad_norm": 1.3280410766601562,
+      "learning_rate": 8.655027167692603e-06,
+      "loss": 0.2029,
+      "step": 19900
+    },
+    {
+      "epoch": 0.26214665473495813,
+      "grad_norm": 1.574950933456421,
+      "learning_rate": 8.652117426144183e-06,
+      "loss": 0.1933,
+      "step": 19920
+    },
+    {
+      "epoch": 0.2624098541875033,
+      "grad_norm": 1.658387303352356,
+      "learning_rate": 8.649205030752995e-06,
+      "loss": 0.2021,
+      "step": 19940
+    },
+    {
+      "epoch": 0.26267305364004845,
+      "grad_norm": 1.4233416318893433,
+      "learning_rate": 8.646289983635354e-06,
+      "loss": 0.1833,
+      "step": 19960
+    },
+    {
+      "epoch": 0.26293625309259355,
+      "grad_norm": 1.3883696794509888,
+      "learning_rate": 8.643372286909505e-06,
+      "loss": 0.205,
+      "step": 19980
+    },
+    {
+      "epoch": 0.2631994525451387,
+      "grad_norm": 1.5209466218948364,
+      "learning_rate": 8.640451942695617e-06,
+      "loss": 0.1945,
+      "step": 20000
+    },
+    {
+      "epoch": 0.2631994525451387,
+      "eval_loss": 0.23008936643600464,
+      "eval_runtime": 3.6892,
+      "eval_samples_per_second": 4.337,
+      "eval_steps_per_second": 2.169,
+      "step": 20000
+    },
+    {
+      "epoch": 0.26346265199768387,
+      "grad_norm": 1.8788349628448486,
+      "learning_rate": 8.637528953115784e-06,
+      "loss": 0.1766,
+      "step": 20020
+    },
+    {
+      "epoch": 0.263725851450229,
+      "grad_norm": 1.1552551984786987,
+      "learning_rate": 8.634603320294022e-06,
+      "loss": 0.2077,
+      "step": 20040
+    },
+    {
+      "epoch": 0.26398905090277414,
+      "grad_norm": 1.2992335557937622,
+      "learning_rate": 8.631675046356266e-06,
+      "loss": 0.1962,
+      "step": 20060
+    },
+    {
+      "epoch": 0.26425225035531924,
+      "grad_norm": 2.216648578643799,
+      "learning_rate": 8.628744133430374e-06,
+      "loss": 0.2024,
+      "step": 20080
+    },
+    {
+      "epoch": 0.2645154498078644,
+      "grad_norm": 2.001193046569824,
+      "learning_rate": 8.625810583646115e-06,
+      "loss": 0.192,
+      "step": 20100
+    },
+    {
+      "epoch": 0.26477864926040956,
+      "grad_norm": 1.3884174823760986,
+      "learning_rate": 8.622874399135183e-06,
+      "loss": 0.2124,
+      "step": 20120
+    },
+    {
+      "epoch": 0.26504184871295466,
+      "grad_norm": 1.5859051942825317,
+      "learning_rate": 8.619935582031181e-06,
+      "loss": 0.2083,
+      "step": 20140
+    },
+    {
+      "epoch": 0.2653050481654998,
+      "grad_norm": 1.4554270505905151,
+      "learning_rate": 8.616994134469622e-06,
+      "loss": 0.2045,
+      "step": 20160
+    },
+    {
+      "epoch": 0.26556824761804493,
+      "grad_norm": 0.9921600222587585,
+      "learning_rate": 8.61405005858794e-06,
+      "loss": 0.2125,
+      "step": 20180
+    },
+    {
+      "epoch": 0.2658314470705901,
+      "grad_norm": 1.617067813873291,
+      "learning_rate": 8.611103356525469e-06,
+      "loss": 0.1888,
+      "step": 20200
+    },
+    {
+      "epoch": 0.26609464652313525,
+      "grad_norm": 1.340311050415039,
+      "learning_rate": 8.60815403042346e-06,
+      "loss": 0.1977,
+      "step": 20220
+    },
+    {
+      "epoch": 0.26635784597568035,
+      "grad_norm": 1.7037678956985474,
+      "learning_rate": 8.60520208242506e-06,
+      "loss": 0.1921,
+      "step": 20240
+    },
+    {
+      "epoch": 0.2666210454282255,
+      "grad_norm": 1.2119412422180176,
+      "learning_rate": 8.602247514675334e-06,
+      "loss": 0.2117,
+      "step": 20260
+    },
+    {
+      "epoch": 0.26688424488077067,
+      "grad_norm": 1.9499931335449219,
+      "learning_rate": 8.59929032932124e-06,
+      "loss": 0.2013,
+      "step": 20280
+    },
+    {
+      "epoch": 0.2671474443333158,
+      "grad_norm": 1.0236115455627441,
+      "learning_rate": 8.596330528511643e-06,
+      "loss": 0.1977,
+      "step": 20300
+    },
+    {
+      "epoch": 0.26741064378586094,
+      "grad_norm": 1.2653539180755615,
+      "learning_rate": 8.59336811439731e-06,
+      "loss": 0.19,
+      "step": 20320
+    },
+    {
+      "epoch": 0.26767384323840604,
+      "grad_norm": 1.5441303253173828,
+      "learning_rate": 8.590403089130903e-06,
+      "loss": 0.2004,
+      "step": 20340
+    },
+    {
+      "epoch": 0.2679370426909512,
+      "grad_norm": 1.658638834953308,
+      "learning_rate": 8.587435454866983e-06,
+      "loss": 0.2096,
+      "step": 20360
+    },
+    {
+      "epoch": 0.26820024214349636,
+      "grad_norm": 1.8121647834777832,
+      "learning_rate": 8.58446521376201e-06,
+      "loss": 0.2186,
+      "step": 20380
+    },
+    {
+      "epoch": 0.26846344159604146,
+      "grad_norm": 1.4915516376495361,
+      "learning_rate": 8.581492367974332e-06,
+      "loss": 0.1757,
+      "step": 20400
+    },
+    {
+      "epoch": 0.2687266410485866,
+      "grad_norm": 1.6920289993286133,
+      "learning_rate": 8.578516919664194e-06,
+      "loss": 0.1944,
+      "step": 20420
+    },
+    {
+      "epoch": 0.2689898405011318,
+      "grad_norm": 1.559740662574768,
+      "learning_rate": 8.575538870993732e-06,
+      "loss": 0.2085,
+      "step": 20440
+    },
+    {
+      "epoch": 0.2692530399536769,
+      "grad_norm": 1.6254700422286987,
+      "learning_rate": 8.572558224126973e-06,
+      "loss": 0.2143,
+      "step": 20460
+    },
+    {
+      "epoch": 0.26951623940622205,
+      "grad_norm": 1.8256478309631348,
+      "learning_rate": 8.569574981229824e-06,
+      "loss": 0.2013,
+      "step": 20480
+    },
+    {
+      "epoch": 0.26977943885876715,
+      "grad_norm": 1.334489107131958,
+      "learning_rate": 8.56658914447009e-06,
+      "loss": 0.1847,
+      "step": 20500
+    },
+    {
+      "epoch": 0.26977943885876715,
+      "eval_loss": 0.23193269968032837,
+      "eval_runtime": 4.0138,
+      "eval_samples_per_second": 3.986,
+      "eval_steps_per_second": 1.993,
+      "step": 20500
+    },
+    {
+      "epoch": 0.2700426383113123,
+      "grad_norm": 1.1153992414474487,
+      "learning_rate": 8.563600716017452e-06,
+      "loss": 0.2105,
+      "step": 20520
+    },
+    {
+      "epoch": 0.27030583776385747,
+      "grad_norm": 1.1996933221817017,
+      "learning_rate": 8.56060969804348e-06,
+      "loss": 0.2314,
+      "step": 20540
+    },
+    {
+      "epoch": 0.2705690372164026,
+      "grad_norm": 1.7793415784835815,
+      "learning_rate": 8.557616092721622e-06,
+      "loss": 0.2004,
+      "step": 20560
+    },
+    {
+      "epoch": 0.27083223666894773,
+      "grad_norm": 1.6759562492370605,
+      "learning_rate": 8.554619902227206e-06,
+      "loss": 0.1966,
+      "step": 20580
+    },
+    {
+      "epoch": 0.2710954361214929,
+      "grad_norm": 1.8329675197601318,
+      "learning_rate": 8.55162112873744e-06,
+      "loss": 0.2135,
+      "step": 20600
+    },
+    {
+      "epoch": 0.271358635574038,
+      "grad_norm": 1.6110550165176392,
+      "learning_rate": 8.548619774431409e-06,
+      "loss": 0.2107,
+      "step": 20620
+    },
+    {
+      "epoch": 0.27162183502658316,
+      "grad_norm": 1.50907301902771,
+      "learning_rate": 8.545615841490075e-06,
+      "loss": 0.189,
+      "step": 20640
+    },
+    {
+      "epoch": 0.27188503447912826,
+      "grad_norm": 1.4752752780914307,
+      "learning_rate": 8.542609332096268e-06,
+      "loss": 0.1896,
+      "step": 20660
+    },
+    {
+      "epoch": 0.2721482339316734,
+      "grad_norm": 1.3202382326126099,
+      "learning_rate": 8.539600248434698e-06,
+      "loss": 0.1884,
+      "step": 20680
+    },
+    {
+      "epoch": 0.2724114333842186,
+      "grad_norm": 1.1638935804367065,
+      "learning_rate": 8.536588592691941e-06,
+      "loss": 0.193,
+      "step": 20700
+    },
+    {
+      "epoch": 0.2726746328367637,
+      "grad_norm": 1.393899917602539,
+      "learning_rate": 8.53357436705644e-06,
+      "loss": 0.1969,
+      "step": 20720
+    },
+    {
+      "epoch": 0.27293783228930885,
+      "grad_norm": 1.9374665021896362,
+      "learning_rate": 8.530557573718512e-06,
+      "loss": 0.1879,
+      "step": 20740
+    },
+    {
+      "epoch": 0.27320103174185395,
+      "grad_norm": 1.219008445739746,
+      "learning_rate": 8.527538214870334e-06,
+      "loss": 0.1804,
+      "step": 20760
+    },
+    {
+      "epoch": 0.2734642311943991,
+      "grad_norm": 1.4541529417037964,
+      "learning_rate": 8.524516292705948e-06,
+      "loss": 0.2107,
+      "step": 20780
+    },
+    {
+      "epoch": 0.27372743064694427,
+      "grad_norm": 1.2216694355010986,
+      "learning_rate": 8.521491809421262e-06,
+      "loss": 0.1805,
+      "step": 20800
+    },
+    {
+      "epoch": 0.2739906300994894,
+      "grad_norm": 0.9781224727630615,
+      "learning_rate": 8.518464767214042e-06,
+      "loss": 0.2154,
+      "step": 20820
+    },
+    {
+      "epoch": 0.27425382955203453,
+      "grad_norm": 0.9742235541343689,
+      "learning_rate": 8.515435168283916e-06,
+      "loss": 0.1905,
+      "step": 20840
+    },
+    {
+      "epoch": 0.2745170290045797,
+      "grad_norm": 1.4652804136276245,
+      "learning_rate": 8.512403014832365e-06,
+      "loss": 0.2211,
+      "step": 20860
+    },
+    {
+      "epoch": 0.2747802284571248,
+      "grad_norm": 1.1766507625579834,
+      "learning_rate": 8.509368309062733e-06,
+      "loss": 0.1857,
+      "step": 20880
+    },
+    {
+      "epoch": 0.27504342790966996,
+      "grad_norm": 1.1451940536499023,
+      "learning_rate": 8.506331053180215e-06,
+      "loss": 0.2087,
+      "step": 20900
+    },
+    {
+      "epoch": 0.27530662736221506,
+      "grad_norm": 1.5707600116729736,
+      "learning_rate": 8.503291249391856e-06,
+      "loss": 0.215,
+      "step": 20920
+    },
+    {
+      "epoch": 0.2755698268147602,
+      "grad_norm": 1.3377816677093506,
+      "learning_rate": 8.500248899906558e-06,
+      "loss": 0.1901,
+      "step": 20940
+    },
+    {
+      "epoch": 0.2758330262673054,
+      "grad_norm": 1.0042674541473389,
+      "learning_rate": 8.49720400693507e-06,
+      "loss": 0.1965,
+      "step": 20960
+    },
+    {
+      "epoch": 0.2760962257198505,
+      "grad_norm": 1.170883059501648,
+      "learning_rate": 8.494156572689993e-06,
+      "loss": 0.1993,
+      "step": 20980
+    },
+    {
+      "epoch": 0.27635942517239565,
+      "grad_norm": 1.0675410032272339,
+      "learning_rate": 8.491106599385766e-06,
+      "loss": 0.1878,
+      "step": 21000
+    },
+    {
+      "epoch": 0.27635942517239565,
+      "eval_loss": 0.22940340638160706,
+      "eval_runtime": 3.3939,
+      "eval_samples_per_second": 4.714,
+      "eval_steps_per_second": 2.357,
+      "step": 21000
+    },
+    {
+      "epoch": 0.2766226246249408,
+      "grad_norm": 1.2435311079025269,
+      "learning_rate": 8.488054089238684e-06,
+      "loss": 0.2104,
+      "step": 21020
+    },
+    {
+      "epoch": 0.2768858240774859,
+      "grad_norm": 1.3752405643463135,
+      "learning_rate": 8.484999044466875e-06,
+      "loss": 0.233,
+      "step": 21040
+    },
+    {
+      "epoch": 0.27714902353003107,
+      "grad_norm": 1.2152206897735596,
+      "learning_rate": 8.481941467290318e-06,
+      "loss": 0.2089,
+      "step": 21060
+    },
+    {
+      "epoch": 0.2774122229825762,
+      "grad_norm": 1.3511422872543335,
+      "learning_rate": 8.478881359930824e-06,
+      "loss": 0.1936,
+      "step": 21080
+    },
+    {
+      "epoch": 0.27767542243512133,
+      "grad_norm": 1.6523436307907104,
+      "learning_rate": 8.47581872461205e-06,
+      "loss": 0.1971,
+      "step": 21100
+    },
+    {
+      "epoch": 0.2779386218876665,
+      "grad_norm": 1.2118182182312012,
+      "learning_rate": 8.472753563559485e-06,
+      "loss": 0.2104,
+      "step": 21120
+    },
+    {
+      "epoch": 0.2782018213402116,
+      "grad_norm": 1.5108622312545776,
+      "learning_rate": 8.469685879000454e-06,
+      "loss": 0.2012,
+      "step": 21140
+    },
+    {
+      "epoch": 0.27846502079275676,
+      "grad_norm": 1.5241140127182007,
+      "learning_rate": 8.46661567316412e-06,
+      "loss": 0.1934,
+      "step": 21160
+    },
+    {
+      "epoch": 0.2787282202453019,
+      "grad_norm": 1.8336952924728394,
+      "learning_rate": 8.46354294828147e-06,
+      "loss": 0.1924,
+      "step": 21180
+    },
+    {
+      "epoch": 0.278991419697847,
+      "grad_norm": 1.293904423713684,
+      "learning_rate": 8.460467706585328e-06,
+      "loss": 0.227,
+      "step": 21200
+    },
+    {
+      "epoch": 0.2792546191503922,
+      "grad_norm": 1.6323716640472412,
+      "learning_rate": 8.457389950310346e-06,
+      "loss": 0.186,
+      "step": 21220
+    },
+    {
+      "epoch": 0.2795178186029373,
+      "grad_norm": 1.5699864625930786,
+      "learning_rate": 8.454309681693003e-06,
+      "loss": 0.2097,
+      "step": 21240
+    },
+    {
+      "epoch": 0.27978101805548244,
+      "grad_norm": 1.8389216661453247,
+      "learning_rate": 8.451226902971599e-06,
+      "loss": 0.2172,
+      "step": 21260
+    },
+    {
+      "epoch": 0.2800442175080276,
+      "grad_norm": 1.5748240947723389,
+      "learning_rate": 8.448141616386265e-06,
+      "loss": 0.2226,
+      "step": 21280
+    },
+    {
+      "epoch": 0.2803074169605727,
+      "grad_norm": 1.8896334171295166,
+      "learning_rate": 8.445053824178951e-06,
+      "loss": 0.1895,
+      "step": 21300
+    },
+    {
+      "epoch": 0.28057061641311787,
+      "grad_norm": 1.423039197921753,
+      "learning_rate": 8.441963528593428e-06,
+      "loss": 0.1944,
+      "step": 21320
+    },
+    {
+      "epoch": 0.280833815865663,
+      "grad_norm": 1.1516410112380981,
+      "learning_rate": 8.438870731875285e-06,
+      "loss": 0.213,
+      "step": 21340
+    },
+    {
+      "epoch": 0.28109701531820813,
+      "grad_norm": 1.0793589353561401,
+      "learning_rate": 8.43577543627193e-06,
+      "loss": 0.1847,
+      "step": 21360
+    },
+    {
+      "epoch": 0.2813602147707533,
+      "grad_norm": 1.4301460981369019,
+      "learning_rate": 8.432677644032586e-06,
+      "loss": 0.1994,
+      "step": 21380
+    },
+    {
+      "epoch": 0.2816234142232984,
+      "grad_norm": 1.631665825843811,
+      "learning_rate": 8.429577357408292e-06,
+      "loss": 0.1962,
+      "step": 21400
+    },
+    {
+      "epoch": 0.28188661367584356,
+      "grad_norm": 0.9818753004074097,
+      "learning_rate": 8.426474578651898e-06,
+      "loss": 0.2016,
+      "step": 21420
+    },
+    {
+      "epoch": 0.2821498131283887,
+      "grad_norm": 1.2682099342346191,
+      "learning_rate": 8.423369310018062e-06,
+      "loss": 0.2102,
+      "step": 21440
+    },
+    {
+      "epoch": 0.2824130125809338,
+      "grad_norm": 1.6712019443511963,
+      "learning_rate": 8.420261553763257e-06,
+      "loss": 0.1859,
+      "step": 21460
+    },
+    {
+      "epoch": 0.282676212033479,
+      "grad_norm": 1.51772940158844,
+      "learning_rate": 8.417151312145762e-06,
+      "loss": 0.1971,
+      "step": 21480
+    },
+    {
+      "epoch": 0.2829394114860241,
+      "grad_norm": 1.2687534093856812,
+      "learning_rate": 8.414038587425658e-06,
+      "loss": 0.1947,
+      "step": 21500
+    },
+    {
+      "epoch": 0.2829394114860241,
+      "eval_loss": 0.23433636128902435,
+      "eval_runtime": 4.0754,
+      "eval_samples_per_second": 3.926,
+      "eval_steps_per_second": 1.963,
+      "step": 21500
+    },
+    {
+      "epoch": 0.28320261093856924,
+      "grad_norm": 1.7680004835128784,
+      "learning_rate": 8.410923381864837e-06,
+      "loss": 0.2159,
+      "step": 21520
+    },
+    {
+      "epoch": 0.2834658103911144,
+      "grad_norm": 2.064258575439453,
+      "learning_rate": 8.407805697726988e-06,
+      "loss": 0.1902,
+      "step": 21540
+    },
+    {
+      "epoch": 0.2837290098436595,
+      "grad_norm": 1.5447962284088135,
+      "learning_rate": 8.404685537277601e-06,
+      "loss": 0.203,
+      "step": 21560
+    },
+    {
+      "epoch": 0.28399220929620467,
+      "grad_norm": 1.272517442703247,
+      "learning_rate": 8.401562902783974e-06,
+      "loss": 0.2192,
+      "step": 21580
+    },
+    {
+      "epoch": 0.2842554087487498,
+      "grad_norm": 1.6980940103530884,
+      "learning_rate": 8.39843779651519e-06,
+      "loss": 0.2007,
+      "step": 21600
+    },
+    {
+      "epoch": 0.28451860820129493,
+      "grad_norm": 1.5552314519882202,
+      "learning_rate": 8.395310220742134e-06,
+      "loss": 0.1896,
+      "step": 21620
+    },
+    {
+      "epoch": 0.2847818076538401,
+      "grad_norm": 1.7980581521987915,
+      "learning_rate": 8.392180177737492e-06,
+      "loss": 0.1875,
+      "step": 21640
+    },
+    {
+      "epoch": 0.2850450071063852,
+      "grad_norm": 1.8983385562896729,
+      "learning_rate": 8.389047669775734e-06,
+      "loss": 0.1961,
+      "step": 21660
+    },
+    {
+      "epoch": 0.28530820655893036,
+      "grad_norm": 1.4371198415756226,
+      "learning_rate": 8.385912699133122e-06,
+      "loss": 0.1972,
+      "step": 21680
+    },
+    {
+      "epoch": 0.2855714060114755,
+      "grad_norm": 1.2548373937606812,
+      "learning_rate": 8.382775268087712e-06,
+      "loss": 0.2248,
+      "step": 21700
+    },
+    {
+      "epoch": 0.2858346054640206,
+      "grad_norm": 2.2749571800231934,
+      "learning_rate": 8.379635378919346e-06,
+      "loss": 0.199,
+      "step": 21720
+    },
+    {
+      "epoch": 0.2860978049165658,
+      "grad_norm": 1.3386566638946533,
+      "learning_rate": 8.37649303390965e-06,
+      "loss": 0.1777,
+      "step": 21740
+    },
+    {
+      "epoch": 0.28636100436911094,
+      "grad_norm": 1.6129246950149536,
+      "learning_rate": 8.373348235342038e-06,
+      "loss": 0.1964,
+      "step": 21760
+    },
+    {
+      "epoch": 0.28662420382165604,
+      "grad_norm": 1.331016182899475,
+      "learning_rate": 8.370200985501702e-06,
+      "loss": 0.1898,
+      "step": 21780
+    },
+    {
+      "epoch": 0.2868874032742012,
+      "grad_norm": 0.9099927544593811,
+      "learning_rate": 8.367051286675622e-06,
+      "loss": 0.1819,
+      "step": 21800
+    },
+    {
+      "epoch": 0.2871506027267463,
+      "grad_norm": 1.780510425567627,
+      "learning_rate": 8.363899141152552e-06,
+      "loss": 0.1989,
+      "step": 21820
+    },
+    {
+      "epoch": 0.28741380217929147,
+      "grad_norm": 1.1993281841278076,
+      "learning_rate": 8.360744551223024e-06,
+      "loss": 0.2135,
+      "step": 21840
+    },
+    {
+      "epoch": 0.2876770016318366,
+      "grad_norm": 1.1239761114120483,
+      "learning_rate": 8.357587519179354e-06,
+      "loss": 0.1926,
+      "step": 21860
+    },
+    {
+      "epoch": 0.28794020108438173,
+      "grad_norm": 1.2809375524520874,
+      "learning_rate": 8.354428047315619e-06,
+      "loss": 0.1904,
+      "step": 21880
+    },
+    {
+      "epoch": 0.2882034005369269,
+      "grad_norm": 1.1730543375015259,
+      "learning_rate": 8.351266137927682e-06,
+      "loss": 0.2111,
+      "step": 21900
+    },
+    {
+      "epoch": 0.288466599989472,
+      "grad_norm": 0.9010506868362427,
+      "learning_rate": 8.34810179331317e-06,
+      "loss": 0.1873,
+      "step": 21920
+    },
+    {
+      "epoch": 0.28872979944201715,
+      "grad_norm": 1.2539600133895874,
+      "learning_rate": 8.344935015771482e-06,
+      "loss": 0.1968,
+      "step": 21940
+    },
+    {
+      "epoch": 0.2889929988945623,
+      "grad_norm": 1.763822078704834,
+      "learning_rate": 8.341765807603783e-06,
+      "loss": 0.1816,
+      "step": 21960
+    },
+    {
+      "epoch": 0.2892561983471074,
+      "grad_norm": 1.3377456665039062,
+      "learning_rate": 8.338594171113009e-06,
+      "loss": 0.2081,
+      "step": 21980
+    },
+    {
+      "epoch": 0.2895193977996526,
+      "grad_norm": 1.2802425622940063,
+      "learning_rate": 8.335420108603854e-06,
+      "loss": 0.2067,
+      "step": 22000
+    },
+    {
+      "epoch": 0.2895193977996526,
+      "eval_loss": 0.22954228520393372,
+      "eval_runtime": 3.8618,
+      "eval_samples_per_second": 4.143,
+      "eval_steps_per_second": 2.072,
+      "step": 22000
+    },
+    {
+      "epoch": 0.28978259725219774,
+      "grad_norm": 1.5020798444747925,
+      "learning_rate": 8.332243622382778e-06,
+      "loss": 0.2019,
+      "step": 22020
+    },
+    {
+      "epoch": 0.29004579670474284,
+      "grad_norm": 1.769721269607544,
+      "learning_rate": 8.329064714758003e-06,
+      "loss": 0.1983,
+      "step": 22040
+    },
+    {
+      "epoch": 0.290308996157288,
+      "grad_norm": 1.4141044616699219,
+      "learning_rate": 8.32588338803951e-06,
+      "loss": 0.2239,
+      "step": 22060
+    },
+    {
+      "epoch": 0.2905721956098331,
+      "grad_norm": 1.486328125,
+      "learning_rate": 8.322699644539036e-06,
+      "loss": 0.2161,
+      "step": 22080
+    },
+    {
+      "epoch": 0.29083539506237827,
+      "grad_norm": 1.4759843349456787,
+      "learning_rate": 8.319513486570078e-06,
+      "loss": 0.2067,
+      "step": 22100
+    },
+    {
+      "epoch": 0.2910985945149234,
+      "grad_norm": 1.5312565565109253,
+      "learning_rate": 8.316324916447886e-06,
+      "loss": 0.2036,
+      "step": 22120
+    },
+    {
+      "epoch": 0.29136179396746853,
+      "grad_norm": 1.4937235116958618,
+      "learning_rate": 8.313133936489461e-06,
+      "loss": 0.1995,
+      "step": 22140
+    },
+    {
+      "epoch": 0.2916249934200137,
+      "grad_norm": 1.7456142902374268,
+      "learning_rate": 8.309940549013555e-06,
+      "loss": 0.194,
+      "step": 22160
+    },
+    {
+      "epoch": 0.29188819287255885,
+      "grad_norm": 1.1423606872558594,
+      "learning_rate": 8.30674475634067e-06,
+      "loss": 0.1954,
+      "step": 22180
+    },
+    {
+      "epoch": 0.29215139232510395,
+      "grad_norm": 1.1642340421676636,
+      "learning_rate": 8.30354656079306e-06,
+      "loss": 0.1915,
+      "step": 22200
+    },
+    {
+      "epoch": 0.2924145917776491,
+      "grad_norm": 1.3926680088043213,
+      "learning_rate": 8.30034596469472e-06,
+      "loss": 0.2152,
+      "step": 22220
+    },
+    {
+      "epoch": 0.2926777912301942,
+      "grad_norm": 1.3315140008926392,
+      "learning_rate": 8.297142970371389e-06,
+      "loss": 0.1889,
+      "step": 22240
+    },
+    {
+      "epoch": 0.2929409906827394,
+      "grad_norm": 1.6384892463684082,
+      "learning_rate": 8.293937580150551e-06,
+      "loss": 0.2019,
+      "step": 22260
+    },
+    {
+      "epoch": 0.29320419013528454,
+      "grad_norm": 1.2769097089767456,
+      "learning_rate": 8.290729796361433e-06,
+      "loss": 0.1894,
+      "step": 22280
+    },
+    {
+      "epoch": 0.29346738958782964,
+      "grad_norm": 1.5622721910476685,
+      "learning_rate": 8.287519621334998e-06,
+      "loss": 0.1969,
+      "step": 22300
+    },
+    {
+      "epoch": 0.2937305890403748,
+      "grad_norm": 1.513412356376648,
+      "learning_rate": 8.284307057403947e-06,
+      "loss": 0.196,
+      "step": 22320
+    },
+    {
+      "epoch": 0.29399378849291996,
+      "grad_norm": 1.8187060356140137,
+      "learning_rate": 8.281092106902715e-06,
+      "loss": 0.1907,
+      "step": 22340
+    },
+    {
+      "epoch": 0.29425698794546506,
+      "grad_norm": 1.238123893737793,
+      "learning_rate": 8.277874772167477e-06,
+      "loss": 0.2122,
+      "step": 22360
+    },
+    {
+      "epoch": 0.2945201873980102,
+      "grad_norm": 1.3014450073242188,
+      "learning_rate": 8.274655055536136e-06,
+      "loss": 0.1975,
+      "step": 22380
+    },
+    {
+      "epoch": 0.29478338685055533,
+      "grad_norm": 1.5750097036361694,
+      "learning_rate": 8.27143295934833e-06,
+      "loss": 0.1947,
+      "step": 22400
+    },
+    {
+      "epoch": 0.2950465863031005,
+      "grad_norm": 1.6815719604492188,
+      "learning_rate": 8.268208485945414e-06,
+      "loss": 0.2004,
+      "step": 22420
+    },
+    {
+      "epoch": 0.29530978575564565,
+      "grad_norm": 1.2699369192123413,
+      "learning_rate": 8.26498163767049e-06,
+      "loss": 0.1982,
+      "step": 22440
+    },
+    {
+      "epoch": 0.29557298520819075,
+      "grad_norm": 1.5934218168258667,
+      "learning_rate": 8.261752416868371e-06,
+      "loss": 0.1881,
+      "step": 22460
+    },
+    {
+      "epoch": 0.2958361846607359,
+      "grad_norm": 1.1055444478988647,
+      "learning_rate": 8.2585208258856e-06,
+      "loss": 0.198,
+      "step": 22480
+    },
+    {
+      "epoch": 0.296099384113281,
+      "grad_norm": 1.6795449256896973,
+      "learning_rate": 8.255286867070435e-06,
+      "loss": 0.1888,
+      "step": 22500
+    },
+    {
+      "epoch": 0.296099384113281,
+      "eval_loss": 0.22831732034683228,
+      "eval_runtime": 4.0235,
+      "eval_samples_per_second": 3.977,
+      "eval_steps_per_second": 1.988,
+      "step": 22500
+    },
+    {
+      "epoch": 0.2963625835658262,
+      "grad_norm": 1.772860050201416,
+      "learning_rate": 8.252050542772865e-06,
+      "loss": 0.213,
+      "step": 22520
+    },
+    {
+      "epoch": 0.29662578301837134,
+      "grad_norm": 1.661737084388733,
+      "learning_rate": 8.248811855344593e-06,
+      "loss": 0.1869,
+      "step": 22540
+    },
+    {
+      "epoch": 0.29688898247091644,
+      "grad_norm": 1.3421753644943237,
+      "learning_rate": 8.245570807139038e-06,
+      "loss": 0.1965,
+      "step": 22560
+    },
+    {
+      "epoch": 0.2971521819234616,
+      "grad_norm": 1.3185909986495972,
+      "learning_rate": 8.242327400511336e-06,
+      "loss": 0.2069,
+      "step": 22580
+    },
+    {
+      "epoch": 0.29741538137600676,
+      "grad_norm": 1.3046684265136719,
+      "learning_rate": 8.239081637818339e-06,
+      "loss": 0.1989,
+      "step": 22600
+    },
+    {
+      "epoch": 0.29767858082855186,
+      "grad_norm": 1.8199553489685059,
+      "learning_rate": 8.235833521418602e-06,
+      "loss": 0.1966,
+      "step": 22620
+    },
+    {
+      "epoch": 0.297941780281097,
+      "grad_norm": 1.520100712776184,
+      "learning_rate": 8.232583053672405e-06,
+      "loss": 0.1817,
+      "step": 22640
+    },
+    {
+      "epoch": 0.29820497973364213,
+      "grad_norm": 2.0245590209960938,
+      "learning_rate": 8.229330236941722e-06,
+      "loss": 0.1855,
+      "step": 22660
+    },
+    {
+      "epoch": 0.2984681791861873,
+      "grad_norm": 1.522176742553711,
+      "learning_rate": 8.226075073590245e-06,
+      "loss": 0.1832,
+      "step": 22680
+    },
+    {
+      "epoch": 0.29873137863873245,
+      "grad_norm": 1.524124026298523,
+      "learning_rate": 8.222817565983365e-06,
+      "loss": 0.1882,
+      "step": 22700
+    },
+    {
+      "epoch": 0.29899457809127755,
+      "grad_norm": 1.4729480743408203,
+      "learning_rate": 8.219557716488178e-06,
+      "loss": 0.1857,
+      "step": 22720
+    },
+    {
+      "epoch": 0.2992577775438227,
+      "grad_norm": 1.5069499015808105,
+      "learning_rate": 8.216295527473482e-06,
+      "loss": 0.2103,
+      "step": 22740
+    },
+    {
+      "epoch": 0.29952097699636787,
+      "grad_norm": 1.5801595449447632,
+      "learning_rate": 8.213031001309773e-06,
+      "loss": 0.1903,
+      "step": 22760
+    },
+    {
+      "epoch": 0.299784176448913,
+      "grad_norm": 1.860127568244934,
+      "learning_rate": 8.20976414036925e-06,
+      "loss": 0.1905,
+      "step": 22780
+    },
+    {
+      "epoch": 0.30004737590145814,
+      "grad_norm": 1.5178112983703613,
+      "learning_rate": 8.206494947025805e-06,
+      "loss": 0.2182,
+      "step": 22800
+    },
+    {
+      "epoch": 0.30031057535400324,
+      "grad_norm": 1.2953382730484009,
+      "learning_rate": 8.203223423655024e-06,
+      "loss": 0.1916,
+      "step": 22820
+    },
+    {
+      "epoch": 0.3005737748065484,
+      "grad_norm": 1.022343635559082,
+      "learning_rate": 8.199949572634191e-06,
+      "loss": 0.1973,
+      "step": 22840
+    },
+    {
+      "epoch": 0.30083697425909356,
+      "grad_norm": 1.61997389793396,
+      "learning_rate": 8.196673396342276e-06,
+      "loss": 0.1876,
+      "step": 22860
+    },
+    {
+      "epoch": 0.30110017371163866,
+      "grad_norm": 1.4742511510849,
+      "learning_rate": 8.193394897159942e-06,
+      "loss": 0.1935,
+      "step": 22880
+    },
+    {
+      "epoch": 0.3013633731641838,
+      "grad_norm": 1.710501790046692,
+      "learning_rate": 8.190114077469538e-06,
+      "loss": 0.2011,
+      "step": 22900
+    },
+    {
+      "epoch": 0.301626572616729,
+      "grad_norm": 1.5145598649978638,
+      "learning_rate": 8.1868309396551e-06,
+      "loss": 0.1843,
+      "step": 22920
+    },
+    {
+      "epoch": 0.3018897720692741,
+      "grad_norm": 1.6268866062164307,
+      "learning_rate": 8.18354548610235e-06,
+      "loss": 0.1863,
+      "step": 22940
+    },
+    {
+      "epoch": 0.30215297152181925,
+      "grad_norm": 1.6191786527633667,
+      "learning_rate": 8.180257719198688e-06,
+      "loss": 0.1929,
+      "step": 22960
+    },
+    {
+      "epoch": 0.30241617097436435,
+      "grad_norm": 1.711717963218689,
+      "learning_rate": 8.176967641333203e-06,
+      "loss": 0.1765,
+      "step": 22980
+    },
+    {
+      "epoch": 0.3026793704269095,
+      "grad_norm": 1.7652232646942139,
+      "learning_rate": 8.173675254896653e-06,
+      "loss": 0.2024,
+      "step": 23000
+    },
+    {
+      "epoch": 0.3026793704269095,
+      "eval_loss": 0.22603467106819153,
+      "eval_runtime": 3.8764,
+      "eval_samples_per_second": 4.128,
+      "eval_steps_per_second": 2.064,
+      "step": 23000
+    },
+    {
+      "epoch": 0.30294256987945467,
+      "grad_norm": 2.284491777420044,
+      "learning_rate": 8.170380562281483e-06,
+      "loss": 0.1788,
+      "step": 23020
+    },
+    {
+      "epoch": 0.3032057693319998,
+      "grad_norm": 1.4521714448928833,
+      "learning_rate": 8.167083565881811e-06,
+      "loss": 0.1922,
+      "step": 23040
+    },
+    {
+      "epoch": 0.30346896878454493,
+      "grad_norm": 1.4119036197662354,
+      "learning_rate": 8.163784268093425e-06,
+      "loss": 0.1963,
+      "step": 23060
+    },
+    {
+      "epoch": 0.3037321682370901,
+      "grad_norm": 1.2874096632003784,
+      "learning_rate": 8.160482671313789e-06,
+      "loss": 0.177,
+      "step": 23080
+    },
+    {
+      "epoch": 0.3039953676896352,
+      "grad_norm": 1.8756635189056396,
+      "learning_rate": 8.157178777942037e-06,
+      "loss": 0.2116,
+      "step": 23100
+    },
+    {
+      "epoch": 0.30425856714218036,
+      "grad_norm": 1.5634640455245972,
+      "learning_rate": 8.153872590378971e-06,
+      "loss": 0.2087,
+      "step": 23120
+    },
+    {
+      "epoch": 0.30452176659472546,
+      "grad_norm": 1.433118224143982,
+      "learning_rate": 8.150564111027064e-06,
+      "loss": 0.189,
+      "step": 23140
+    },
+    {
+      "epoch": 0.3047849660472706,
+      "grad_norm": 1.481082797050476,
+      "learning_rate": 8.14725334229045e-06,
+      "loss": 0.2028,
+      "step": 23160
+    },
+    {
+      "epoch": 0.3050481654998158,
+      "grad_norm": 1.3750746250152588,
+      "learning_rate": 8.143940286574923e-06,
+      "loss": 0.2014,
+      "step": 23180
+    },
+    {
+      "epoch": 0.3053113649523609,
+      "grad_norm": 1.3214579820632935,
+      "learning_rate": 8.14062494628795e-06,
+      "loss": 0.1908,
+      "step": 23200
+    },
+    {
+      "epoch": 0.30557456440490605,
+      "grad_norm": 1.325783371925354,
+      "learning_rate": 8.13730732383865e-06,
+      "loss": 0.207,
+      "step": 23220
+    },
+    {
+      "epoch": 0.30583776385745115,
+      "grad_norm": 1.6514558792114258,
+      "learning_rate": 8.133987421637798e-06,
+      "loss": 0.1937,
+      "step": 23240
+    },
+    {
+      "epoch": 0.3061009633099963,
+      "grad_norm": 1.625745415687561,
+      "learning_rate": 8.130665242097836e-06,
+      "loss": 0.216,
+      "step": 23260
+    },
+    {
+      "epoch": 0.30636416276254147,
+      "grad_norm": 1.286608099937439,
+      "learning_rate": 8.127340787632851e-06,
+      "loss": 0.1838,
+      "step": 23280
+    },
+    {
+      "epoch": 0.3066273622150866,
+      "grad_norm": 1.647371768951416,
+      "learning_rate": 8.124014060658588e-06,
+      "loss": 0.1794,
+      "step": 23300
+    },
+    {
+      "epoch": 0.30689056166763173,
+      "grad_norm": 1.6688404083251953,
+      "learning_rate": 8.120685063592439e-06,
+      "loss": 0.2182,
+      "step": 23320
+    },
+    {
+      "epoch": 0.3071537611201769,
+      "grad_norm": 1.4828112125396729,
+      "learning_rate": 8.117353798853452e-06,
+      "loss": 0.1977,
+      "step": 23340
+    },
+    {
+      "epoch": 0.307416960572722,
+      "grad_norm": 1.3157685995101929,
+      "learning_rate": 8.114020268862318e-06,
+      "loss": 0.2013,
+      "step": 23360
+    },
+    {
+      "epoch": 0.30768016002526716,
+      "grad_norm": 1.348103404045105,
+      "learning_rate": 8.110684476041374e-06,
+      "loss": 0.1958,
+      "step": 23380
+    },
+    {
+      "epoch": 0.30794335947781226,
+      "grad_norm": 1.3627320528030396,
+      "learning_rate": 8.107346422814609e-06,
+      "loss": 0.1964,
+      "step": 23400
+    },
+    {
+      "epoch": 0.3082065589303574,
+      "grad_norm": 1.2069445848464966,
+      "learning_rate": 8.10400611160764e-06,
+      "loss": 0.1842,
+      "step": 23420
+    },
+    {
+      "epoch": 0.3084697583829026,
+      "grad_norm": 1.1975723505020142,
+      "learning_rate": 8.100663544847737e-06,
+      "loss": 0.1989,
+      "step": 23440
+    },
+    {
+      "epoch": 0.3087329578354477,
+      "grad_norm": 0.9581167697906494,
+      "learning_rate": 8.097318724963806e-06,
+      "loss": 0.196,
+      "step": 23460
+    },
+    {
+      "epoch": 0.30899615728799285,
+      "grad_norm": 1.351913571357727,
+      "learning_rate": 8.093971654386386e-06,
+      "loss": 0.1971,
+      "step": 23480
+    },
+    {
+      "epoch": 0.309259356740538,
+      "grad_norm": 1.1794575452804565,
+      "learning_rate": 8.090622335547657e-06,
+      "loss": 0.2159,
+      "step": 23500
+    },
+    {
+      "epoch": 0.309259356740538,
+      "eval_loss": 0.23287689685821533,
+      "eval_runtime": 4.084,
+      "eval_samples_per_second": 3.918,
+      "eval_steps_per_second": 1.959,
+      "step": 23500
+    },
+    {
+      "epoch": 0.3095225561930831,
+      "grad_norm": 1.535636067390442,
+      "learning_rate": 8.087270770881434e-06,
+      "loss": 0.1854,
+      "step": 23520
+    },
+    {
+      "epoch": 0.30978575564562827,
+      "grad_norm": 1.610330581665039,
+      "learning_rate": 8.083916962823154e-06,
+      "loss": 0.1969,
+      "step": 23540
+    },
+    {
+      "epoch": 0.3100489550981734,
+      "grad_norm": 1.2447261810302734,
+      "learning_rate": 8.080560913809893e-06,
+      "loss": 0.1971,
+      "step": 23560
+    },
+    {
+      "epoch": 0.31031215455071853,
+      "grad_norm": 1.936592698097229,
+      "learning_rate": 8.077202626280356e-06,
+      "loss": 0.1859,
+      "step": 23580
+    },
+    {
+      "epoch": 0.3105753540032637,
+      "grad_norm": 1.390691876411438,
+      "learning_rate": 8.073842102674871e-06,
+      "loss": 0.193,
+      "step": 23600
+    },
+    {
+      "epoch": 0.3108385534558088,
+      "grad_norm": 1.335228681564331,
+      "learning_rate": 8.070479345435387e-06,
+      "loss": 0.191,
+      "step": 23620
+    },
+    {
+      "epoch": 0.31110175290835396,
+      "grad_norm": 1.8836966753005981,
+      "learning_rate": 8.067114357005486e-06,
+      "loss": 0.1949,
+      "step": 23640
+    },
+    {
+      "epoch": 0.3113649523608991,
+      "grad_norm": 1.7667146921157837,
+      "learning_rate": 8.063747139830363e-06,
+      "loss": 0.1912,
+      "step": 23660
+    },
+    {
+      "epoch": 0.3116281518134442,
+      "grad_norm": 1.1574208736419678,
+      "learning_rate": 8.060377696356837e-06,
+      "loss": 0.1948,
+      "step": 23680
+    },
+    {
+      "epoch": 0.3118913512659894,
+      "grad_norm": 1.4678550958633423,
+      "learning_rate": 8.05700602903334e-06,
+      "loss": 0.1961,
+      "step": 23700
+    },
+    {
+      "epoch": 0.3121545507185345,
+      "grad_norm": 1.9705437421798706,
+      "learning_rate": 8.053632140309928e-06,
+      "loss": 0.193,
+      "step": 23720
+    },
+    {
+      "epoch": 0.31241775017107964,
+      "grad_norm": 1.0036622285842896,
+      "learning_rate": 8.050256032638264e-06,
+      "loss": 0.1853,
+      "step": 23740
+    },
+    {
+      "epoch": 0.3126809496236248,
+      "grad_norm": 1.4243701696395874,
+      "learning_rate": 8.046877708471625e-06,
+      "loss": 0.2066,
+      "step": 23760
+    },
+    {
+      "epoch": 0.3129441490761699,
+      "grad_norm": 1.387059211730957,
+      "learning_rate": 8.043497170264903e-06,
+      "loss": 0.2003,
+      "step": 23780
+    },
+    {
+      "epoch": 0.31320734852871507,
+      "grad_norm": 1.5255814790725708,
+      "learning_rate": 8.040114420474591e-06,
+      "loss": 0.2024,
+      "step": 23800
+    },
+    {
+      "epoch": 0.3134705479812602,
+      "grad_norm": 1.2056115865707397,
+      "learning_rate": 8.036729461558798e-06,
+      "loss": 0.1954,
+      "step": 23820
+    },
+    {
+      "epoch": 0.31373374743380533,
+      "grad_norm": 1.7188072204589844,
+      "learning_rate": 8.033342295977232e-06,
+      "loss": 0.2067,
+      "step": 23840
+    },
+    {
+      "epoch": 0.3139969468863505,
+      "grad_norm": 1.562418818473816,
+      "learning_rate": 8.029952926191203e-06,
+      "loss": 0.1936,
+      "step": 23860
+    },
+    {
+      "epoch": 0.3142601463388956,
+      "grad_norm": 1.6976962089538574,
+      "learning_rate": 8.026561354663629e-06,
+      "loss": 0.1934,
+      "step": 23880
+    },
+    {
+      "epoch": 0.31452334579144076,
+      "grad_norm": 1.4061962366104126,
+      "learning_rate": 8.023167583859026e-06,
+      "loss": 0.1904,
+      "step": 23900
+    },
+    {
+      "epoch": 0.3147865452439859,
+      "grad_norm": 1.5870304107666016,
+      "learning_rate": 8.019771616243503e-06,
+      "loss": 0.2017,
+      "step": 23920
+    },
+    {
+      "epoch": 0.315049744696531,
+      "grad_norm": 1.471400260925293,
+      "learning_rate": 8.016373454284772e-06,
+      "loss": 0.1987,
+      "step": 23940
+    },
+    {
+      "epoch": 0.3153129441490762,
+      "grad_norm": 1.7155166864395142,
+      "learning_rate": 8.01297310045214e-06,
+      "loss": 0.2009,
+      "step": 23960
+    },
+    {
+      "epoch": 0.3155761436016213,
+      "grad_norm": 1.6650073528289795,
+      "learning_rate": 8.009570557216494e-06,
+      "loss": 0.1861,
+      "step": 23980
+    },
+    {
+      "epoch": 0.31583934305416644,
+      "grad_norm": 1.2807759046554565,
+      "learning_rate": 8.006165827050331e-06,
+      "loss": 0.2037,
+      "step": 24000
+    },
+    {
+      "epoch": 0.31583934305416644,
+      "eval_loss": 0.2269577980041504,
+      "eval_runtime": 3.5894,
+      "eval_samples_per_second": 4.458,
+      "eval_steps_per_second": 2.229,
+      "step": 24000
+    },
+    {
+      "epoch": 0.3161025425067116,
+      "grad_norm": 1.3005039691925049,
+      "learning_rate": 8.002758912427724e-06,
+      "loss": 0.193,
+      "step": 24020
+    },
+    {
+      "epoch": 0.3163657419592567,
+      "grad_norm": 1.6803559064865112,
+      "learning_rate": 7.999349815824337e-06,
+      "loss": 0.1981,
+      "step": 24040
+    },
+    {
+      "epoch": 0.31662894141180187,
+      "grad_norm": 1.61122465133667,
+      "learning_rate": 7.99593853971742e-06,
+      "loss": 0.1929,
+      "step": 24060
+    },
+    {
+      "epoch": 0.316892140864347,
+      "grad_norm": 1.5391247272491455,
+      "learning_rate": 7.992525086585807e-06,
+      "loss": 0.2055,
+      "step": 24080
+    },
+    {
+      "epoch": 0.31715534031689213,
+      "grad_norm": 1.3271325826644897,
+      "learning_rate": 7.989109458909912e-06,
+      "loss": 0.1891,
+      "step": 24100
+    },
+    {
+      "epoch": 0.3174185397694373,
+      "grad_norm": 1.4182229042053223,
+      "learning_rate": 7.98569165917173e-06,
+      "loss": 0.196,
+      "step": 24120
+    },
+    {
+      "epoch": 0.3176817392219824,
+      "grad_norm": 2.0993266105651855,
+      "learning_rate": 7.982271689854834e-06,
+      "loss": 0.2078,
+      "step": 24140
+    },
+    {
+      "epoch": 0.31794493867452756,
+      "grad_norm": 1.3305984735488892,
+      "learning_rate": 7.978849553444378e-06,
+      "loss": 0.2075,
+      "step": 24160
+    },
+    {
+      "epoch": 0.3182081381270727,
+      "grad_norm": 1.3937575817108154,
+      "learning_rate": 7.975425252427085e-06,
+      "loss": 0.2104,
+      "step": 24180
+    },
+    {
+      "epoch": 0.3184713375796178,
+      "grad_norm": 1.5269880294799805,
+      "learning_rate": 7.971998789291254e-06,
+      "loss": 0.2001,
+      "step": 24200
+    },
+    {
+      "epoch": 0.318734537032163,
+      "grad_norm": 1.826543927192688,
+      "learning_rate": 7.968570166526752e-06,
+      "loss": 0.1862,
+      "step": 24220
+    },
+    {
+      "epoch": 0.31899773648470814,
+      "grad_norm": 1.4873958826065063,
+      "learning_rate": 7.965139386625019e-06,
+      "loss": 0.1995,
+      "step": 24240
+    },
+    {
+      "epoch": 0.31926093593725324,
+      "grad_norm": 1.4426296949386597,
+      "learning_rate": 7.961706452079062e-06,
+      "loss": 0.1732,
+      "step": 24260
+    },
+    {
+      "epoch": 0.3195241353897984,
+      "grad_norm": 1.0942785739898682,
+      "learning_rate": 7.958271365383447e-06,
+      "loss": 0.2027,
+      "step": 24280
+    },
+    {
+      "epoch": 0.3197873348423435,
+      "grad_norm": 1.4700877666473389,
+      "learning_rate": 7.954834129034318e-06,
+      "loss": 0.1915,
+      "step": 24300
+    },
+    {
+      "epoch": 0.32005053429488867,
+      "grad_norm": 1.8832855224609375,
+      "learning_rate": 7.951394745529366e-06,
+      "loss": 0.1811,
+      "step": 24320
+    },
+    {
+      "epoch": 0.3203137337474338,
+      "grad_norm": 1.728117823600769,
+      "learning_rate": 7.947953217367852e-06,
+      "loss": 0.2051,
+      "step": 24340
+    },
+    {
+      "epoch": 0.32057693319997893,
+      "grad_norm": 1.547445297241211,
+      "learning_rate": 7.944509547050592e-06,
+      "loss": 0.1842,
+      "step": 24360
+    },
+    {
+      "epoch": 0.3208401326525241,
+      "grad_norm": 1.4591413736343384,
+      "learning_rate": 7.941063737079958e-06,
+      "loss": 0.1826,
+      "step": 24380
+    },
+    {
+      "epoch": 0.3211033321050692,
+      "grad_norm": 1.6257224082946777,
+      "learning_rate": 7.937615789959877e-06,
+      "loss": 0.2123,
+      "step": 24400
+    },
+    {
+      "epoch": 0.32136653155761435,
+      "grad_norm": 1.4552556276321411,
+      "learning_rate": 7.93416570819583e-06,
+      "loss": 0.205,
+      "step": 24420
+    },
+    {
+      "epoch": 0.3216297310101595,
+      "grad_norm": 1.6078482866287231,
+      "learning_rate": 7.930713494294849e-06,
+      "loss": 0.1949,
+      "step": 24440
+    },
+    {
+      "epoch": 0.3218929304627046,
+      "grad_norm": 1.637121319770813,
+      "learning_rate": 7.927259150765518e-06,
+      "loss": 0.1998,
+      "step": 24460
+    },
+    {
+      "epoch": 0.3221561299152498,
+      "grad_norm": 1.4053151607513428,
+      "learning_rate": 7.923802680117962e-06,
+      "loss": 0.2,
+      "step": 24480
+    },
+    {
+      "epoch": 0.32241932936779494,
+      "grad_norm": 1.6685831546783447,
+      "learning_rate": 7.920344084863856e-06,
+      "loss": 0.1943,
+      "step": 24500
+    },
+    {
+      "epoch": 0.32241932936779494,
+      "eval_loss": 0.2286447286605835,
+      "eval_runtime": 3.7745,
+      "eval_samples_per_second": 4.239,
+      "eval_steps_per_second": 2.12,
+      "step": 24500
+    },
+    {
+      "epoch": 0.32268252882034004,
+      "grad_norm": 1.7281101942062378,
+      "learning_rate": 7.916883367516417e-06,
+      "loss": 0.1908,
+      "step": 24520
+    },
+    {
+      "epoch": 0.3229457282728852,
+      "grad_norm": 1.1933457851409912,
+      "learning_rate": 7.913420530590409e-06,
+      "loss": 0.2072,
+      "step": 24540
+    },
+    {
+      "epoch": 0.3232089277254303,
+      "grad_norm": 1.7660480737686157,
+      "learning_rate": 7.909955576602131e-06,
+      "loss": 0.1968,
+      "step": 24560
+    },
+    {
+      "epoch": 0.32347212717797547,
+      "grad_norm": 1.7296861410140991,
+      "learning_rate": 7.90648850806942e-06,
+      "loss": 0.1889,
+      "step": 24580
+    },
+    {
+      "epoch": 0.3237353266305206,
+      "grad_norm": 1.403404951095581,
+      "learning_rate": 7.903019327511654e-06,
+      "loss": 0.1942,
+      "step": 24600
+    },
+    {
+      "epoch": 0.32399852608306573,
+      "grad_norm": 1.1963797807693481,
+      "learning_rate": 7.899548037449741e-06,
+      "loss": 0.1932,
+      "step": 24620
+    },
+    {
+      "epoch": 0.3242617255356109,
+      "grad_norm": 1.5024147033691406,
+      "learning_rate": 7.896074640406125e-06,
+      "loss": 0.2018,
+      "step": 24640
+    },
+    {
+      "epoch": 0.32452492498815605,
+      "grad_norm": 1.5994799137115479,
+      "learning_rate": 7.892599138904779e-06,
+      "loss": 0.1824,
+      "step": 24660
+    },
+    {
+      "epoch": 0.32478812444070115,
+      "grad_norm": 1.3419272899627686,
+      "learning_rate": 7.88912153547121e-06,
+      "loss": 0.1894,
+      "step": 24680
+    },
+    {
+      "epoch": 0.3250513238932463,
+      "grad_norm": 1.2723745107650757,
+      "learning_rate": 7.885641832632442e-06,
+      "loss": 0.1909,
+      "step": 24700
+    },
+    {
+      "epoch": 0.3253145233457914,
+      "grad_norm": 1.8302538394927979,
+      "learning_rate": 7.882160032917036e-06,
+      "loss": 0.1963,
+      "step": 24720
+    },
+    {
+      "epoch": 0.3255777227983366,
+      "grad_norm": 1.5681018829345703,
+      "learning_rate": 7.878676138855072e-06,
+      "loss": 0.1974,
+      "step": 24740
+    },
+    {
+      "epoch": 0.32584092225088174,
+      "grad_norm": 1.1832281351089478,
+      "learning_rate": 7.87519015297815e-06,
+      "loss": 0.1818,
+      "step": 24760
+    },
+    {
+      "epoch": 0.32610412170342684,
+      "grad_norm": 1.5395869016647339,
+      "learning_rate": 7.871702077819392e-06,
+      "loss": 0.1999,
+      "step": 24780
+    },
+    {
+      "epoch": 0.326367321155972,
+      "grad_norm": 1.3865679502487183,
+      "learning_rate": 7.868211915913437e-06,
+      "loss": 0.2071,
+      "step": 24800
+    },
+    {
+      "epoch": 0.32663052060851716,
+      "grad_norm": 1.1856433153152466,
+      "learning_rate": 7.864719669796446e-06,
+      "loss": 0.1852,
+      "step": 24820
+    },
+    {
+      "epoch": 0.32689372006106227,
+      "grad_norm": 1.5135220289230347,
+      "learning_rate": 7.861225342006087e-06,
+      "loss": 0.1939,
+      "step": 24840
+    },
+    {
+      "epoch": 0.3271569195136074,
+      "grad_norm": 1.6115458011627197,
+      "learning_rate": 7.857728935081542e-06,
+      "loss": 0.1955,
+      "step": 24860
+    },
+    {
+      "epoch": 0.32742011896615253,
+      "grad_norm": 1.3511749505996704,
+      "learning_rate": 7.854230451563505e-06,
+      "loss": 0.2018,
+      "step": 24880
+    },
+    {
+      "epoch": 0.3276833184186977,
+      "grad_norm": 1.3730037212371826,
+      "learning_rate": 7.850729893994184e-06,
+      "loss": 0.2154,
+      "step": 24900
+    },
+    {
+      "epoch": 0.32794651787124285,
+      "grad_norm": 1.7922899723052979,
+      "learning_rate": 7.847227264917285e-06,
+      "loss": 0.1974,
+      "step": 24920
+    },
+    {
+      "epoch": 0.32820971732378795,
+      "grad_norm": 1.5185973644256592,
+      "learning_rate": 7.843722566878027e-06,
+      "loss": 0.1908,
+      "step": 24940
+    },
+    {
+      "epoch": 0.3284729167763331,
+      "grad_norm": 1.9613237380981445,
+      "learning_rate": 7.840215802423126e-06,
+      "loss": 0.1955,
+      "step": 24960
+    },
+    {
+      "epoch": 0.3287361162288782,
+      "grad_norm": 0.8838825821876526,
+      "learning_rate": 7.836706974100805e-06,
+      "loss": 0.2182,
+      "step": 24980
+    },
+    {
+      "epoch": 0.3289993156814234,
+      "grad_norm": 1.4697967767715454,
+      "learning_rate": 7.833196084460786e-06,
+      "loss": 0.1802,
+      "step": 25000
+    },
+    {
+      "epoch": 0.3289993156814234,
+      "eval_loss": 0.2300737202167511,
+      "eval_runtime": 4.1975,
+      "eval_samples_per_second": 3.812,
+      "eval_steps_per_second": 1.906,
+      "step": 25000
+    },
+    {
+      "epoch": 0.32926251513396854,
+      "grad_norm": 2.1109511852264404,
+      "learning_rate": 7.829683136054282e-06,
+      "loss": 0.2025,
+      "step": 25020
+    },
+    {
+      "epoch": 0.32952571458651364,
+      "grad_norm": 1.1279417276382446,
+      "learning_rate": 7.826168131434011e-06,
+      "loss": 0.1943,
+      "step": 25040
+    },
+    {
+      "epoch": 0.3297889140390588,
+      "grad_norm": 1.1264922618865967,
+      "learning_rate": 7.822651073154183e-06,
+      "loss": 0.2004,
+      "step": 25060
+    },
+    {
+      "epoch": 0.33005211349160396,
+      "grad_norm": 1.4738670587539673,
+      "learning_rate": 7.819131963770497e-06,
+      "loss": 0.1862,
+      "step": 25080
+    },
+    {
+      "epoch": 0.33031531294414906,
+      "grad_norm": 1.3449814319610596,
+      "learning_rate": 7.815610805840143e-06,
+      "loss": 0.1994,
+      "step": 25100
+    },
+    {
+      "epoch": 0.3305785123966942,
+      "grad_norm": 1.5339714288711548,
+      "learning_rate": 7.812087601921803e-06,
+      "loss": 0.1922,
+      "step": 25120
+    },
+    {
+      "epoch": 0.33084171184923933,
+      "grad_norm": 1.3457015752792358,
+      "learning_rate": 7.808562354575642e-06,
+      "loss": 0.1973,
+      "step": 25140
+    },
+    {
+      "epoch": 0.3311049113017845,
+      "grad_norm": 1.417813777923584,
+      "learning_rate": 7.805035066363313e-06,
+      "loss": 0.2065,
+      "step": 25160
+    },
+    {
+      "epoch": 0.33136811075432965,
+      "grad_norm": 2.1004390716552734,
+      "learning_rate": 7.801505739847951e-06,
+      "loss": 0.186,
+      "step": 25180
+    },
+    {
+      "epoch": 0.33163131020687475,
+      "grad_norm": 1.7762999534606934,
+      "learning_rate": 7.79797437759417e-06,
+      "loss": 0.1946,
+      "step": 25200
+    },
+    {
+      "epoch": 0.3318945096594199,
+      "grad_norm": 1.1182222366333008,
+      "learning_rate": 7.794440982168066e-06,
+      "loss": 0.2139,
+      "step": 25220
+    },
+    {
+      "epoch": 0.33215770911196507,
+      "grad_norm": 1.0348156690597534,
+      "learning_rate": 7.790905556137212e-06,
+      "loss": 0.1858,
+      "step": 25240
+    },
+    {
+      "epoch": 0.3324209085645102,
+      "grad_norm": 1.7715905904769897,
+      "learning_rate": 7.787368102070653e-06,
+      "loss": 0.1949,
+      "step": 25260
+    },
+    {
+      "epoch": 0.33268410801705534,
+      "grad_norm": 1.2989587783813477,
+      "learning_rate": 7.783828622538914e-06,
+      "loss": 0.1891,
+      "step": 25280
+    },
+    {
+      "epoch": 0.33294730746960044,
+      "grad_norm": 1.4975996017456055,
+      "learning_rate": 7.780287120113987e-06,
+      "loss": 0.1778,
+      "step": 25300
+    },
+    {
+      "epoch": 0.3332105069221456,
+      "grad_norm": 1.1541359424591064,
+      "learning_rate": 7.776743597369336e-06,
+      "loss": 0.1946,
+      "step": 25320
+    },
+    {
+      "epoch": 0.33347370637469076,
+      "grad_norm": 1.0933555364608765,
+      "learning_rate": 7.773198056879892e-06,
+      "loss": 0.2042,
+      "step": 25340
+    },
+    {
+      "epoch": 0.33373690582723586,
+      "grad_norm": 1.4401134252548218,
+      "learning_rate": 7.769650501222052e-06,
+      "loss": 0.2196,
+      "step": 25360
+    },
+    {
+      "epoch": 0.334000105279781,
+      "grad_norm": 1.6898517608642578,
+      "learning_rate": 7.76610093297368e-06,
+      "loss": 0.1933,
+      "step": 25380
+    },
+    {
+      "epoch": 0.3342633047323262,
+      "grad_norm": 1.2631891965866089,
+      "learning_rate": 7.762549354714098e-06,
+      "loss": 0.2011,
+      "step": 25400
+    },
+    {
+      "epoch": 0.3345265041848713,
+      "grad_norm": 1.3703383207321167,
+      "learning_rate": 7.758995769024094e-06,
+      "loss": 0.1982,
+      "step": 25420
+    },
+    {
+      "epoch": 0.33478970363741645,
+      "grad_norm": 1.7224010229110718,
+      "learning_rate": 7.755440178485908e-06,
+      "loss": 0.1943,
+      "step": 25440
+    },
+    {
+      "epoch": 0.33505290308996155,
+      "grad_norm": 1.2067534923553467,
+      "learning_rate": 7.751882585683246e-06,
+      "loss": 0.1965,
+      "step": 25460
+    },
+    {
+      "epoch": 0.3353161025425067,
+      "grad_norm": 1.271046757698059,
+      "learning_rate": 7.748322993201259e-06,
+      "loss": 0.1976,
+      "step": 25480
+    },
+    {
+      "epoch": 0.33557930199505187,
+      "grad_norm": 1.5270521640777588,
+      "learning_rate": 7.744761403626557e-06,
+      "loss": 0.1858,
+      "step": 25500
+    },
+    {
+      "epoch": 0.33557930199505187,
+      "eval_loss": 0.22894707322120667,
+      "eval_runtime": 3.6736,
+      "eval_samples_per_second": 4.355,
+      "eval_steps_per_second": 2.178,
+      "step": 25500
+    },
+    {
+      "epoch": 0.335842501447597,
+      "grad_norm": 1.4875749349594116,
+      "learning_rate": 7.741197819547202e-06,
+      "loss": 0.201,
+      "step": 25520
+    },
+    {
+      "epoch": 0.33610570090014213,
+      "grad_norm": 1.5831831693649292,
+      "learning_rate": 7.7376322435527e-06,
+      "loss": 0.1994,
+      "step": 25540
+    },
+    {
+      "epoch": 0.33636890035268724,
+      "grad_norm": 1.4568507671356201,
+      "learning_rate": 7.73406467823401e-06,
+      "loss": 0.2021,
+      "step": 25560
+    },
+    {
+      "epoch": 0.3366320998052324,
+      "grad_norm": 1.7448711395263672,
+      "learning_rate": 7.730495126183533e-06,
+      "loss": 0.2105,
+      "step": 25580
+    },
+    {
+      "epoch": 0.33689529925777756,
+      "grad_norm": 1.5118181705474854,
+      "learning_rate": 7.726923589995114e-06,
+      "loss": 0.1905,
+      "step": 25600
+    },
+    {
+      "epoch": 0.33715849871032266,
+      "grad_norm": 1.7153593301773071,
+      "learning_rate": 7.723350072264046e-06,
+      "loss": 0.1961,
+      "step": 25620
+    },
+    {
+      "epoch": 0.3374216981628678,
+      "grad_norm": 1.8910669088363647,
+      "learning_rate": 7.71977457558705e-06,
+      "loss": 0.1896,
+      "step": 25640
+    },
+    {
+      "epoch": 0.337684897615413,
+      "grad_norm": 1.4960875511169434,
+      "learning_rate": 7.716197102562296e-06,
+      "loss": 0.1999,
+      "step": 25660
+    },
+    {
+      "epoch": 0.3379480970679581,
+      "grad_norm": 1.0270928144454956,
+      "learning_rate": 7.71261765578938e-06,
+      "loss": 0.2,
+      "step": 25680
+    },
+    {
+      "epoch": 0.33821129652050325,
+      "grad_norm": 1.4495872259140015,
+      "learning_rate": 7.709036237869347e-06,
+      "loss": 0.2037,
+      "step": 25700
+    },
+    {
+      "epoch": 0.33847449597304835,
+      "grad_norm": 1.598889708518982,
+      "learning_rate": 7.705452851404658e-06,
+      "loss": 0.1987,
+      "step": 25720
+    },
+    {
+      "epoch": 0.3387376954255935,
+      "grad_norm": 1.6163400411605835,
+      "learning_rate": 7.701867498999215e-06,
+      "loss": 0.1912,
+      "step": 25740
+    },
+    {
+      "epoch": 0.33900089487813867,
+      "grad_norm": 1.6725608110427856,
+      "learning_rate": 7.698280183258344e-06,
+      "loss": 0.2066,
+      "step": 25760
+    },
+    {
+      "epoch": 0.3392640943306838,
+      "grad_norm": 1.0591260194778442,
+      "learning_rate": 7.694690906788801e-06,
+      "loss": 0.1911,
+      "step": 25780
+    },
+    {
+      "epoch": 0.33952729378322893,
+      "grad_norm": 1.868281364440918,
+      "learning_rate": 7.691099672198763e-06,
+      "loss": 0.1844,
+      "step": 25800
+    },
+    {
+      "epoch": 0.3397904932357741,
+      "grad_norm": 1.9380199909210205,
+      "learning_rate": 7.687506482097832e-06,
+      "loss": 0.1851,
+      "step": 25820
+    },
+    {
+      "epoch": 0.3400536926883192,
+      "grad_norm": 1.061417579650879,
+      "learning_rate": 7.683911339097033e-06,
+      "loss": 0.22,
+      "step": 25840
+    },
+    {
+      "epoch": 0.34031689214086436,
+      "grad_norm": 1.226642370223999,
+      "learning_rate": 7.680314245808804e-06,
+      "loss": 0.2029,
+      "step": 25860
+    },
+    {
+      "epoch": 0.34058009159340946,
+      "grad_norm": 1.2876148223876953,
+      "learning_rate": 7.676715204847005e-06,
+      "loss": 0.1979,
+      "step": 25880
+    },
+    {
+      "epoch": 0.3408432910459546,
+      "grad_norm": 1.2829890251159668,
+      "learning_rate": 7.673114218826912e-06,
+      "loss": 0.1907,
+      "step": 25900
+    },
+    {
+      "epoch": 0.3411064904984998,
+      "grad_norm": 1.5666148662567139,
+      "learning_rate": 7.669511290365212e-06,
+      "loss": 0.2028,
+      "step": 25920
+    },
+    {
+      "epoch": 0.3413696899510449,
+      "grad_norm": 1.81540048122406,
+      "learning_rate": 7.665906422080004e-06,
+      "loss": 0.1946,
+      "step": 25940
+    },
+    {
+      "epoch": 0.34163288940359005,
+      "grad_norm": 1.497800588607788,
+      "learning_rate": 7.662299616590798e-06,
+      "loss": 0.1955,
+      "step": 25960
+    },
+    {
+      "epoch": 0.3418960888561352,
+      "grad_norm": 1.3465317487716675,
+      "learning_rate": 7.658690876518509e-06,
+      "loss": 0.1768,
+      "step": 25980
+    },
+    {
+      "epoch": 0.3421592883086803,
+      "grad_norm": 1.4544472694396973,
+      "learning_rate": 7.655080204485459e-06,
+      "loss": 0.1999,
+      "step": 26000
+    },
+    {
+      "epoch": 0.3421592883086803,
+      "eval_loss": 0.23119953274726868,
+      "eval_runtime": 3.9285,
+      "eval_samples_per_second": 4.073,
+      "eval_steps_per_second": 2.036,
+      "step": 26000
+    },
+    {
+      "epoch": 0.34242248776122547,
+      "grad_norm": 1.9522703886032104,
+      "learning_rate": 7.651467603115374e-06,
+      "loss": 0.1861,
+      "step": 26020
+    },
+    {
+      "epoch": 0.3426856872137706,
+      "grad_norm": 1.3395721912384033,
+      "learning_rate": 7.647853075033385e-06,
+      "loss": 0.1859,
+      "step": 26040
+    },
+    {
+      "epoch": 0.34294888666631573,
+      "grad_norm": 1.8510384559631348,
+      "learning_rate": 7.644236622866015e-06,
+      "loss": 0.1853,
+      "step": 26060
+    },
+    {
+      "epoch": 0.3432120861188609,
+      "grad_norm": 1.4481117725372314,
+      "learning_rate": 7.640618249241195e-06,
+      "loss": 0.2006,
+      "step": 26080
+    },
+    {
+      "epoch": 0.343475285571406,
+      "grad_norm": 2.429398775100708,
+      "learning_rate": 7.636997956788244e-06,
+      "loss": 0.1989,
+      "step": 26100
+    },
+    {
+      "epoch": 0.34373848502395116,
+      "grad_norm": 1.6355116367340088,
+      "learning_rate": 7.633375748137882e-06,
+      "loss": 0.215,
+      "step": 26120
+    },
+    {
+      "epoch": 0.34400168447649626,
+      "grad_norm": 1.4148768186569214,
+      "learning_rate": 7.629751625922217e-06,
+      "loss": 0.1651,
+      "step": 26140
+    },
+    {
+      "epoch": 0.3442648839290414,
+      "grad_norm": 1.441811442375183,
+      "learning_rate": 7.626125592774749e-06,
+      "loss": 0.1932,
+      "step": 26160
+    },
+    {
+      "epoch": 0.3445280833815866,
+      "grad_norm": 1.583533525466919,
+      "learning_rate": 7.622497651330365e-06,
+      "loss": 0.186,
+      "step": 26180
+    },
+    {
+      "epoch": 0.3447912828341317,
+      "grad_norm": 1.115140438079834,
+      "learning_rate": 7.618867804225342e-06,
+      "loss": 0.2083,
+      "step": 26200
+    },
+    {
+      "epoch": 0.34505448228667684,
+      "grad_norm": 1.6831294298171997,
+      "learning_rate": 7.615236054097339e-06,
+      "loss": 0.1934,
+      "step": 26220
+    },
+    {
+      "epoch": 0.345317681739222,
+      "grad_norm": 1.7468053102493286,
+      "learning_rate": 7.6116024035853985e-06,
+      "loss": 0.1957,
+      "step": 26240
+    },
+    {
+      "epoch": 0.3455808811917671,
+      "grad_norm": 1.554460883140564,
+      "learning_rate": 7.6079668553299435e-06,
+      "loss": 0.187,
+      "step": 26260
+    },
+    {
+      "epoch": 0.34584408064431227,
+      "grad_norm": 2.365882396697998,
+      "learning_rate": 7.604329411972779e-06,
+      "loss": 0.2031,
+      "step": 26280
+    },
+    {
+      "epoch": 0.3461072800968574,
+      "grad_norm": 1.2920475006103516,
+      "learning_rate": 7.600690076157081e-06,
+      "loss": 0.1951,
+      "step": 26300
+    },
+    {
+      "epoch": 0.34637047954940253,
+      "grad_norm": 1.9068281650543213,
+      "learning_rate": 7.597048850527409e-06,
+      "loss": 0.1806,
+      "step": 26320
+    },
+    {
+      "epoch": 0.3466336790019477,
+      "grad_norm": 1.191720724105835,
+      "learning_rate": 7.5934057377296875e-06,
+      "loss": 0.2108,
+      "step": 26340
+    },
+    {
+      "epoch": 0.3468968784544928,
+      "grad_norm": 1.8883830308914185,
+      "learning_rate": 7.5897607404112184e-06,
+      "loss": 0.2083,
+      "step": 26360
+    },
+    {
+      "epoch": 0.34716007790703796,
+      "grad_norm": 1.1917402744293213,
+      "learning_rate": 7.586113861220668e-06,
+      "loss": 0.1909,
+      "step": 26380
+    },
+    {
+      "epoch": 0.3474232773595831,
+      "grad_norm": 1.2995685338974,
+      "learning_rate": 7.582465102808075e-06,
+      "loss": 0.1886,
+      "step": 26400
+    },
+    {
+      "epoch": 0.3476864768121282,
+      "grad_norm": 1.4324181079864502,
+      "learning_rate": 7.57881446782484e-06,
+      "loss": 0.1916,
+      "step": 26420
+    },
+    {
+      "epoch": 0.3479496762646734,
+      "grad_norm": 1.7305119037628174,
+      "learning_rate": 7.5751619589237294e-06,
+      "loss": 0.1989,
+      "step": 26440
+    },
+    {
+      "epoch": 0.3482128757172185,
+      "grad_norm": 1.4263875484466553,
+      "learning_rate": 7.57150757875887e-06,
+      "loss": 0.179,
+      "step": 26460
+    },
+    {
+      "epoch": 0.34847607516976364,
+      "grad_norm": 1.6204602718353271,
+      "learning_rate": 7.567851329985749e-06,
+      "loss": 0.169,
+      "step": 26480
+    },
+    {
+      "epoch": 0.3487392746223088,
+      "grad_norm": 1.4380024671554565,
+      "learning_rate": 7.5641932152612105e-06,
+      "loss": 0.1982,
+      "step": 26500
+    },
+    {
+      "epoch": 0.3487392746223088,
+      "eval_loss": 0.22670982778072357,
+      "eval_runtime": 3.9742,
+      "eval_samples_per_second": 4.026,
+      "eval_steps_per_second": 2.013,
+      "step": 26500
+    },
+    {
+      "epoch": 0.3490024740748539,
+      "grad_norm": 2.1273958683013916,
+      "learning_rate": 7.560533237243455e-06,
+      "loss": 0.1881,
+      "step": 26520
+    },
+    {
+      "epoch": 0.34926567352739907,
+      "grad_norm": 1.4395248889923096,
+      "learning_rate": 7.556871398592041e-06,
+      "loss": 0.1862,
+      "step": 26540
+    },
+    {
+      "epoch": 0.3495288729799442,
+      "grad_norm": 1.1848291158676147,
+      "learning_rate": 7.553207701967869e-06,
+      "loss": 0.1931,
+      "step": 26560
+    },
+    {
+      "epoch": 0.34979207243248933,
+      "grad_norm": 1.5863664150238037,
+      "learning_rate": 7.549542150033201e-06,
+      "loss": 0.1935,
+      "step": 26580
+    },
+    {
+      "epoch": 0.3500552718850345,
+      "grad_norm": 1.4963582754135132,
+      "learning_rate": 7.54587474545164e-06,
+      "loss": 0.1965,
+      "step": 26600
+    },
+    {
+      "epoch": 0.3503184713375796,
+      "grad_norm": 1.4116308689117432,
+      "learning_rate": 7.542205490888137e-06,
+      "loss": 0.1947,
+      "step": 26620
+    },
+    {
+      "epoch": 0.35058167079012476,
+      "grad_norm": 1.6497915983200073,
+      "learning_rate": 7.538534389008989e-06,
+      "loss": 0.1785,
+      "step": 26640
+    },
+    {
+      "epoch": 0.3508448702426699,
+      "grad_norm": 1.1521989107131958,
+      "learning_rate": 7.534861442481833e-06,
+      "loss": 0.1911,
+      "step": 26660
+    },
+    {
+      "epoch": 0.351108069695215,
+      "grad_norm": 1.6722307205200195,
+      "learning_rate": 7.531186653975649e-06,
+      "loss": 0.1974,
+      "step": 26680
+    },
+    {
+      "epoch": 0.3513712691477602,
+      "grad_norm": 1.4373106956481934,
+      "learning_rate": 7.527510026160749e-06,
+      "loss": 0.2048,
+      "step": 26700
+    },
+    {
+      "epoch": 0.35163446860030534,
+      "grad_norm": 1.6447272300720215,
+      "learning_rate": 7.523831561708792e-06,
+      "loss": 0.1842,
+      "step": 26720
+    },
+    {
+      "epoch": 0.35189766805285044,
+      "grad_norm": 1.451189398765564,
+      "learning_rate": 7.5201512632927654e-06,
+      "loss": 0.1808,
+      "step": 26740
+    },
+    {
+      "epoch": 0.3521608675053956,
+      "grad_norm": 1.3760697841644287,
+      "learning_rate": 7.516469133586988e-06,
+      "loss": 0.2041,
+      "step": 26760
+    },
+    {
+      "epoch": 0.3524240669579407,
+      "grad_norm": 1.388441562652588,
+      "learning_rate": 7.512785175267111e-06,
+      "loss": 0.2063,
+      "step": 26780
+    },
+    {
+      "epoch": 0.35268726641048587,
+      "grad_norm": 1.7355960607528687,
+      "learning_rate": 7.509099391010115e-06,
+      "loss": 0.1993,
+      "step": 26800
+    },
+    {
+      "epoch": 0.352950465863031,
+      "grad_norm": 1.8026609420776367,
+      "learning_rate": 7.505411783494308e-06,
+      "loss": 0.1946,
+      "step": 26820
+    },
+    {
+      "epoch": 0.35321366531557613,
+      "grad_norm": 1.534008264541626,
+      "learning_rate": 7.501722355399322e-06,
+      "loss": 0.1912,
+      "step": 26840
+    },
+    {
+      "epoch": 0.3534768647681213,
+      "grad_norm": 0.95521479845047,
+      "learning_rate": 7.498031109406111e-06,
+      "loss": 0.1859,
+      "step": 26860
+    },
+    {
+      "epoch": 0.3537400642206664,
+      "grad_norm": 1.5450291633605957,
+      "learning_rate": 7.494338048196952e-06,
+      "loss": 0.1733,
+      "step": 26880
+    },
+    {
+      "epoch": 0.35400326367321155,
+      "grad_norm": 1.8953852653503418,
+      "learning_rate": 7.490643174455439e-06,
+      "loss": 0.201,
+      "step": 26900
+    },
+    {
+      "epoch": 0.3542664631257567,
+      "grad_norm": 1.484449863433838,
+      "learning_rate": 7.4869464908664835e-06,
+      "loss": 0.1736,
+      "step": 26920
+    },
+    {
+      "epoch": 0.3545296625783018,
+      "grad_norm": 1.2346879243850708,
+      "learning_rate": 7.483248000116313e-06,
+      "loss": 0.2004,
+      "step": 26940
+    },
+    {
+      "epoch": 0.354792862030847,
+      "grad_norm": 1.7133169174194336,
+      "learning_rate": 7.479547704892469e-06,
+      "loss": 0.195,
+      "step": 26960
+    },
+    {
+      "epoch": 0.35505606148339214,
+      "grad_norm": 1.655983328819275,
+      "learning_rate": 7.475845607883801e-06,
+      "loss": 0.1749,
+      "step": 26980
+    },
+    {
+      "epoch": 0.35531926093593724,
+      "grad_norm": 1.578018307685852,
+      "learning_rate": 7.472141711780474e-06,
+      "loss": 0.2048,
+      "step": 27000
+    },
+    {
+      "epoch": 0.35531926093593724,
+      "eval_loss": 0.2275880128145218,
+      "eval_runtime": 3.7013,
+      "eval_samples_per_second": 4.323,
+      "eval_steps_per_second": 2.161,
+      "step": 27000
+    },
+    {
+      "epoch": 0.3555824603884824,
+      "grad_norm": 1.000169038772583,
+      "learning_rate": 7.4684360192739515e-06,
+      "loss": 0.2025,
+      "step": 27020
+    },
+    {
+      "epoch": 0.3558456598410275,
+      "grad_norm": 1.5644018650054932,
+      "learning_rate": 7.4647285330570095e-06,
+      "loss": 0.1898,
+      "step": 27040
+    },
+    {
+      "epoch": 0.35610885929357267,
+      "grad_norm": 1.7656346559524536,
+      "learning_rate": 7.461019255823723e-06,
+      "loss": 0.2014,
+      "step": 27060
+    },
+    {
+      "epoch": 0.3563720587461178,
+      "grad_norm": 1.3626211881637573,
+      "learning_rate": 7.457308190269473e-06,
+      "loss": 0.2036,
+      "step": 27080
+    },
+    {
+      "epoch": 0.35663525819866293,
+      "grad_norm": 2.267908811569214,
+      "learning_rate": 7.453595339090937e-06,
+      "loss": 0.1874,
+      "step": 27100
+    },
+    {
+      "epoch": 0.3568984576512081,
+      "grad_norm": 1.917323350906372,
+      "learning_rate": 7.4498807049860886e-06,
+      "loss": 0.1753,
+      "step": 27120
+    },
+    {
+      "epoch": 0.35716165710375325,
+      "grad_norm": 1.5948214530944824,
+      "learning_rate": 7.4461642906542e-06,
+      "loss": 0.1965,
+      "step": 27140
+    },
+    {
+      "epoch": 0.35742485655629835,
+      "grad_norm": 1.2101593017578125,
+      "learning_rate": 7.442446098795837e-06,
+      "loss": 0.1878,
+      "step": 27160
+    },
+    {
+      "epoch": 0.3576880560088435,
+      "grad_norm": 1.6462680101394653,
+      "learning_rate": 7.438726132112854e-06,
+      "loss": 0.1951,
+      "step": 27180
+    },
+    {
+      "epoch": 0.3579512554613886,
+      "grad_norm": 1.1773737668991089,
+      "learning_rate": 7.435004393308398e-06,
+      "loss": 0.1988,
+      "step": 27200
+    },
+    {
+      "epoch": 0.3582144549139338,
+      "grad_norm": 1.9109054803848267,
+      "learning_rate": 7.4312808850869025e-06,
+      "loss": 0.1799,
+      "step": 27220
+    },
+    {
+      "epoch": 0.35847765436647894,
+      "grad_norm": 1.7749842405319214,
+      "learning_rate": 7.427555610154087e-06,
+      "loss": 0.1803,
+      "step": 27240
+    },
+    {
+      "epoch": 0.35874085381902404,
+      "grad_norm": 1.4601877927780151,
+      "learning_rate": 7.4238285712169535e-06,
+      "loss": 0.2053,
+      "step": 27260
+    },
+    {
+      "epoch": 0.3590040532715692,
+      "grad_norm": 1.544498085975647,
+      "learning_rate": 7.420099770983792e-06,
+      "loss": 0.2132,
+      "step": 27280
+    },
+    {
+      "epoch": 0.35926725272411436,
+      "grad_norm": 1.5508592128753662,
+      "learning_rate": 7.4163692121641605e-06,
+      "loss": 0.2006,
+      "step": 27300
+    },
+    {
+      "epoch": 0.35953045217665947,
+      "grad_norm": 1.4120047092437744,
+      "learning_rate": 7.412636897468908e-06,
+      "loss": 0.1871,
+      "step": 27320
+    },
+    {
+      "epoch": 0.3597936516292046,
+      "grad_norm": 1.5071078538894653,
+      "learning_rate": 7.4089028296101495e-06,
+      "loss": 0.1933,
+      "step": 27340
+    },
+    {
+      "epoch": 0.36005685108174973,
+      "grad_norm": 1.7866566181182861,
+      "learning_rate": 7.405167011301278e-06,
+      "loss": 0.1975,
+      "step": 27360
+    },
+    {
+      "epoch": 0.3603200505342949,
+      "grad_norm": 1.5075924396514893,
+      "learning_rate": 7.401429445256961e-06,
+      "loss": 0.1924,
+      "step": 27380
+    },
+    {
+      "epoch": 0.36058324998684005,
+      "grad_norm": 1.8227300643920898,
+      "learning_rate": 7.3976901341931315e-06,
+      "loss": 0.1999,
+      "step": 27400
+    },
+    {
+      "epoch": 0.36084644943938515,
+      "grad_norm": 1.5395746231079102,
+      "learning_rate": 7.393949080826991e-06,
+      "loss": 0.1961,
+      "step": 27420
+    },
+    {
+      "epoch": 0.3611096488919303,
+      "grad_norm": 1.919358253479004,
+      "learning_rate": 7.390206287877011e-06,
+      "loss": 0.1782,
+      "step": 27440
+    },
+    {
+      "epoch": 0.3613728483444754,
+      "grad_norm": 1.553295612335205,
+      "learning_rate": 7.386461758062922e-06,
+      "loss": 0.1947,
+      "step": 27460
+    },
+    {
+      "epoch": 0.3616360477970206,
+      "grad_norm": 1.3424113988876343,
+      "learning_rate": 7.38271549410572e-06,
+      "loss": 0.1719,
+      "step": 27480
+    },
+    {
+      "epoch": 0.36189924724956574,
+      "grad_norm": 1.6891669034957886,
+      "learning_rate": 7.37896749872766e-06,
+      "loss": 0.1889,
+      "step": 27500
+    },
+    {
+      "epoch": 0.36189924724956574,
+      "eval_loss": 0.2287963479757309,
+      "eval_runtime": 4.0523,
+      "eval_samples_per_second": 3.948,
+      "eval_steps_per_second": 1.974,
+      "step": 27500
+    },
+    {
+      "epoch": 0.36216244670211084,
+      "grad_norm": 1.382899284362793,
+      "learning_rate": 7.3752177746522565e-06,
+      "loss": 0.1964,
+      "step": 27520
+    },
+    {
+      "epoch": 0.362425646154656,
+      "grad_norm": 1.5320451259613037,
+      "learning_rate": 7.371466324604278e-06,
+      "loss": 0.211,
+      "step": 27540
+    },
+    {
+      "epoch": 0.36268884560720116,
+      "grad_norm": 1.694749355316162,
+      "learning_rate": 7.367713151309748e-06,
+      "loss": 0.1827,
+      "step": 27560
+    },
+    {
+      "epoch": 0.36295204505974626,
+      "grad_norm": 1.4547884464263916,
+      "learning_rate": 7.363958257495943e-06,
+      "loss": 0.1854,
+      "step": 27580
+    },
+    {
+      "epoch": 0.3632152445122914,
+      "grad_norm": 1.615562081336975,
+      "learning_rate": 7.360201645891387e-06,
+      "loss": 0.1826,
+      "step": 27600
+    },
+    {
+      "epoch": 0.36347844396483653,
+      "grad_norm": 1.945508360862732,
+      "learning_rate": 7.3564433192258585e-06,
+      "loss": 0.1912,
+      "step": 27620
+    },
+    {
+      "epoch": 0.3637416434173817,
+      "grad_norm": 1.0960791110992432,
+      "learning_rate": 7.352683280230374e-06,
+      "loss": 0.1929,
+      "step": 27640
+    },
+    {
+      "epoch": 0.36400484286992685,
+      "grad_norm": 0.9601831436157227,
+      "learning_rate": 7.348921531637201e-06,
+      "loss": 0.1795,
+      "step": 27660
+    },
+    {
+      "epoch": 0.36426804232247195,
+      "grad_norm": 1.9509992599487305,
+      "learning_rate": 7.345158076179845e-06,
+      "loss": 0.1923,
+      "step": 27680
+    },
+    {
+      "epoch": 0.3645312417750171,
+      "grad_norm": 1.1728651523590088,
+      "learning_rate": 7.341392916593055e-06,
+      "loss": 0.1877,
+      "step": 27700
+    },
+    {
+      "epoch": 0.36479444122756227,
+      "grad_norm": 1.5046495199203491,
+      "learning_rate": 7.3376260556128165e-06,
+      "loss": 0.1853,
+      "step": 27720
+    },
+    {
+      "epoch": 0.3650576406801074,
+      "grad_norm": 1.2125834226608276,
+      "learning_rate": 7.333857495976351e-06,
+      "loss": 0.1964,
+      "step": 27740
+    },
+    {
+      "epoch": 0.36532084013265254,
+      "grad_norm": 2.138115167617798,
+      "learning_rate": 7.3300872404221144e-06,
+      "loss": 0.2054,
+      "step": 27760
+    },
+    {
+      "epoch": 0.36558403958519764,
+      "grad_norm": 1.9705768823623657,
+      "learning_rate": 7.326315291689798e-06,
+      "loss": 0.2117,
+      "step": 27780
+    },
+    {
+      "epoch": 0.3658472390377428,
+      "grad_norm": 1.3419169187545776,
+      "learning_rate": 7.3225416525203185e-06,
+      "loss": 0.1948,
+      "step": 27800
+    },
+    {
+      "epoch": 0.36611043849028796,
+      "grad_norm": 1.583864450454712,
+      "learning_rate": 7.318766325655826e-06,
+      "loss": 0.1937,
+      "step": 27820
+    },
+    {
+      "epoch": 0.36637363794283306,
+      "grad_norm": 1.134717345237732,
+      "learning_rate": 7.314989313839691e-06,
+      "loss": 0.1838,
+      "step": 27840
+    },
+    {
+      "epoch": 0.3666368373953782,
+      "grad_norm": 1.7151679992675781,
+      "learning_rate": 7.311210619816515e-06,
+      "loss": 0.2038,
+      "step": 27860
+    },
+    {
+      "epoch": 0.3669000368479234,
+      "grad_norm": 1.3344742059707642,
+      "learning_rate": 7.307430246332119e-06,
+      "loss": 0.2036,
+      "step": 27880
+    },
+    {
+      "epoch": 0.3671632363004685,
+      "grad_norm": 1.4910918474197388,
+      "learning_rate": 7.303648196133541e-06,
+      "loss": 0.2211,
+      "step": 27900
+    },
+    {
+      "epoch": 0.36742643575301365,
+      "grad_norm": 1.4586677551269531,
+      "learning_rate": 7.299864471969043e-06,
+      "loss": 0.199,
+      "step": 27920
+    },
+    {
+      "epoch": 0.36768963520555875,
+      "grad_norm": 1.5334227085113525,
+      "learning_rate": 7.2960790765880994e-06,
+      "loss": 0.2058,
+      "step": 27940
+    },
+    {
+      "epoch": 0.3679528346581039,
+      "grad_norm": 1.706739902496338,
+      "learning_rate": 7.292292012741402e-06,
+      "loss": 0.2117,
+      "step": 27960
+    },
+    {
+      "epoch": 0.36821603411064907,
+      "grad_norm": 1.4768829345703125,
+      "learning_rate": 7.2885032831808525e-06,
+      "loss": 0.1994,
+      "step": 27980
+    },
+    {
+      "epoch": 0.3684792335631942,
+      "grad_norm": 1.357080101966858,
+      "learning_rate": 7.284712890659566e-06,
+      "loss": 0.2015,
+      "step": 28000
+    },
+    {
+      "epoch": 0.3684792335631942,
+      "eval_loss": 0.22829023003578186,
+      "eval_runtime": 3.8653,
+      "eval_samples_per_second": 4.139,
+      "eval_steps_per_second": 2.07,
+      "step": 28000
+    },
+    {
+      "epoch": 0.36874243301573933,
+      "grad_norm": 1.650801181793213,
+      "learning_rate": 7.280920837931862e-06,
+      "loss": 0.1871,
+      "step": 28020
+    },
+    {
+      "epoch": 0.36900563246828444,
+      "grad_norm": 1.5779048204421997,
+      "learning_rate": 7.277127127753267e-06,
+      "loss": 0.1933,
+      "step": 28040
+    },
+    {
+      "epoch": 0.3692688319208296,
+      "grad_norm": 1.660475730895996,
+      "learning_rate": 7.273331762880519e-06,
+      "loss": 0.1958,
+      "step": 28060
+    },
+    {
+      "epoch": 0.36953203137337476,
+      "grad_norm": 1.8706557750701904,
+      "learning_rate": 7.269534746071547e-06,
+      "loss": 0.2056,
+      "step": 28080
+    },
+    {
+      "epoch": 0.36979523082591986,
+      "grad_norm": 1.5429123640060425,
+      "learning_rate": 7.265736080085488e-06,
+      "loss": 0.1891,
+      "step": 28100
+    },
+    {
+      "epoch": 0.370058430278465,
+      "grad_norm": 1.2483011484146118,
+      "learning_rate": 7.261935767682678e-06,
+      "loss": 0.1926,
+      "step": 28120
+    },
+    {
+      "epoch": 0.3703216297310102,
+      "grad_norm": 1.6058788299560547,
+      "learning_rate": 7.258133811624643e-06,
+      "loss": 0.1961,
+      "step": 28140
+    },
+    {
+      "epoch": 0.3705848291835553,
+      "grad_norm": 1.7268517017364502,
+      "learning_rate": 7.254330214674111e-06,
+      "loss": 0.1965,
+      "step": 28160
+    },
+    {
+      "epoch": 0.37084802863610045,
+      "grad_norm": 1.4688665866851807,
+      "learning_rate": 7.250524979594995e-06,
+      "loss": 0.1932,
+      "step": 28180
+    },
+    {
+      "epoch": 0.37111122808864555,
+      "grad_norm": 1.3922450542449951,
+      "learning_rate": 7.246718109152405e-06,
+      "loss": 0.196,
+      "step": 28200
+    },
+    {
+      "epoch": 0.3713744275411907,
+      "grad_norm": 1.4392485618591309,
+      "learning_rate": 7.242909606112635e-06,
+      "loss": 0.1938,
+      "step": 28220
+    },
+    {
+      "epoch": 0.37163762699373587,
+      "grad_norm": 1.2701090574264526,
+      "learning_rate": 7.239099473243167e-06,
+      "loss": 0.2191,
+      "step": 28240
+    },
+    {
+      "epoch": 0.371900826446281,
+      "grad_norm": 1.9006272554397583,
+      "learning_rate": 7.235287713312667e-06,
+      "loss": 0.1914,
+      "step": 28260
+    },
+    {
+      "epoch": 0.37216402589882613,
+      "grad_norm": 1.257494330406189,
+      "learning_rate": 7.231474329090984e-06,
+      "loss": 0.1976,
+      "step": 28280
+    },
+    {
+      "epoch": 0.3724272253513713,
+      "grad_norm": 1.5651761293411255,
+      "learning_rate": 7.2276593233491455e-06,
+      "loss": 0.2041,
+      "step": 28300
+    },
+    {
+      "epoch": 0.3726904248039164,
+      "grad_norm": 1.5762892961502075,
+      "learning_rate": 7.2238426988593605e-06,
+      "loss": 0.1986,
+      "step": 28320
+    },
+    {
+      "epoch": 0.37295362425646156,
+      "grad_norm": 1.3466367721557617,
+      "learning_rate": 7.220024458395009e-06,
+      "loss": 0.1956,
+      "step": 28340
+    },
+    {
+      "epoch": 0.37321682370900666,
+      "grad_norm": 1.4380916357040405,
+      "learning_rate": 7.21620460473065e-06,
+      "loss": 0.1819,
+      "step": 28360
+    },
+    {
+      "epoch": 0.3734800231615518,
+      "grad_norm": 1.9113876819610596,
+      "learning_rate": 7.2123831406420146e-06,
+      "loss": 0.1896,
+      "step": 28380
+    },
+    {
+      "epoch": 0.373743222614097,
+      "grad_norm": 2.2138962745666504,
+      "learning_rate": 7.208560068906002e-06,
+      "loss": 0.1995,
+      "step": 28400
+    },
+    {
+      "epoch": 0.3740064220666421,
+      "grad_norm": 1.407368779182434,
+      "learning_rate": 7.204735392300681e-06,
+      "loss": 0.189,
+      "step": 28420
+    },
+    {
+      "epoch": 0.37426962151918725,
+      "grad_norm": 1.3954256772994995,
+      "learning_rate": 7.200909113605284e-06,
+      "loss": 0.2109,
+      "step": 28440
+    },
+    {
+      "epoch": 0.3745328209717324,
+      "grad_norm": 1.079060435295105,
+      "learning_rate": 7.197081235600211e-06,
+      "loss": 0.2058,
+      "step": 28460
+    },
+    {
+      "epoch": 0.3747960204242775,
+      "grad_norm": 1.0610185861587524,
+      "learning_rate": 7.193251761067022e-06,
+      "loss": 0.1892,
+      "step": 28480
+    },
+    {
+      "epoch": 0.37505921987682267,
+      "grad_norm": 1.647186517715454,
+      "learning_rate": 7.1894206927884405e-06,
+      "loss": 0.2047,
+      "step": 28500
+    },
+    {
+      "epoch": 0.37505921987682267,
+      "eval_loss": 0.2245418280363083,
+      "eval_runtime": 3.9708,
+      "eval_samples_per_second": 4.029,
+      "eval_steps_per_second": 2.015,
+      "step": 28500
+    },
+    {
+      "epoch": 0.3753224193293678,
+      "grad_norm": 1.5495457649230957,
+      "learning_rate": 7.1855880335483416e-06,
+      "loss": 0.1748,
+      "step": 28520
+    },
+    {
+      "epoch": 0.37558561878191293,
+      "grad_norm": 1.0980201959609985,
+      "learning_rate": 7.181753786131762e-06,
+      "loss": 0.1847,
+      "step": 28540
+    },
+    {
+      "epoch": 0.3758488182344581,
+      "grad_norm": 1.4752447605133057,
+      "learning_rate": 7.177917953324892e-06,
+      "loss": 0.2025,
+      "step": 28560
+    },
+    {
+      "epoch": 0.3761120176870032,
+      "grad_norm": 1.9669016599655151,
+      "learning_rate": 7.174080537915071e-06,
+      "loss": 0.1694,
+      "step": 28580
+    },
+    {
+      "epoch": 0.37637521713954836,
+      "grad_norm": 2.1316587924957275,
+      "learning_rate": 7.17024154269079e-06,
+      "loss": 0.1919,
+      "step": 28600
+    },
+    {
+      "epoch": 0.37663841659209346,
+      "grad_norm": 1.6368564367294312,
+      "learning_rate": 7.166400970441689e-06,
+      "loss": 0.2134,
+      "step": 28620
+    },
+    {
+      "epoch": 0.3769016160446386,
+      "grad_norm": 1.3808722496032715,
+      "learning_rate": 7.162558823958553e-06,
+      "loss": 0.1935,
+      "step": 28640
+    },
+    {
+      "epoch": 0.3771648154971838,
+      "grad_norm": 1.5966193675994873,
+      "learning_rate": 7.158715106033312e-06,
+      "loss": 0.1839,
+      "step": 28660
+    },
+    {
+      "epoch": 0.3774280149497289,
+      "grad_norm": 1.3168615102767944,
+      "learning_rate": 7.154869819459036e-06,
+      "loss": 0.199,
+      "step": 28680
+    },
+    {
+      "epoch": 0.37769121440227404,
+      "grad_norm": 2.016068935394287,
+      "learning_rate": 7.151022967029936e-06,
+      "loss": 0.2095,
+      "step": 28700
+    },
+    {
+      "epoch": 0.3779544138548192,
+      "grad_norm": 1.157910943031311,
+      "learning_rate": 7.147174551541358e-06,
+      "loss": 0.1822,
+      "step": 28720
+    },
+    {
+      "epoch": 0.3782176133073643,
+      "grad_norm": 1.3194897174835205,
+      "learning_rate": 7.143324575789792e-06,
+      "loss": 0.1881,
+      "step": 28740
+    },
+    {
+      "epoch": 0.37848081275990947,
+      "grad_norm": 1.3568211793899536,
+      "learning_rate": 7.139473042572852e-06,
+      "loss": 0.1837,
+      "step": 28760
+    },
+    {
+      "epoch": 0.3787440122124546,
+      "grad_norm": 1.5873463153839111,
+      "learning_rate": 7.135619954689289e-06,
+      "loss": 0.1907,
+      "step": 28780
+    },
+    {
+      "epoch": 0.37900721166499973,
+      "grad_norm": 1.6537472009658813,
+      "learning_rate": 7.131765314938982e-06,
+      "loss": 0.1881,
+      "step": 28800
+    },
+    {
+      "epoch": 0.3792704111175449,
+      "grad_norm": 1.684956669807434,
+      "learning_rate": 7.12790912612294e-06,
+      "loss": 0.1901,
+      "step": 28820
+    },
+    {
+      "epoch": 0.37953361057009,
+      "grad_norm": 1.281388759613037,
+      "learning_rate": 7.124051391043294e-06,
+      "loss": 0.2202,
+      "step": 28840
+    },
+    {
+      "epoch": 0.37979681002263516,
+      "grad_norm": 1.15740966796875,
+      "learning_rate": 7.120192112503304e-06,
+      "loss": 0.1845,
+      "step": 28860
+    },
+    {
+      "epoch": 0.3800600094751803,
+      "grad_norm": 1.3437304496765137,
+      "learning_rate": 7.116331293307343e-06,
+      "loss": 0.194,
+      "step": 28880
+    },
+    {
+      "epoch": 0.3803232089277254,
+      "grad_norm": 1.584691047668457,
+      "learning_rate": 7.112468936260914e-06,
+      "loss": 0.1971,
+      "step": 28900
+    },
+    {
+      "epoch": 0.3805864083802706,
+      "grad_norm": 1.637677550315857,
+      "learning_rate": 7.10860504417063e-06,
+      "loss": 0.193,
+      "step": 28920
+    },
+    {
+      "epoch": 0.3808496078328157,
+      "grad_norm": 1.6480554342269897,
+      "learning_rate": 7.104739619844222e-06,
+      "loss": 0.1765,
+      "step": 28940
+    },
+    {
+      "epoch": 0.38111280728536084,
+      "grad_norm": 2.04034161567688,
+      "learning_rate": 7.1008726660905315e-06,
+      "loss": 0.1931,
+      "step": 28960
+    },
+    {
+      "epoch": 0.381376006737906,
+      "grad_norm": 1.660254716873169,
+      "learning_rate": 7.097004185719518e-06,
+      "loss": 0.1911,
+      "step": 28980
+    },
+    {
+      "epoch": 0.3816392061904511,
+      "grad_norm": 1.1696051359176636,
+      "learning_rate": 7.093134181542243e-06,
+      "loss": 0.1894,
+      "step": 29000
+    },
+    {
+      "epoch": 0.3816392061904511,
+      "eval_loss": 0.22716061770915985,
+      "eval_runtime": 3.4013,
+      "eval_samples_per_second": 4.704,
+      "eval_steps_per_second": 2.352,
+      "step": 29000
+    },
+    {
+      "epoch": 0.38190240564299627,
+      "grad_norm": 1.4298220872879028,
+      "learning_rate": 7.089262656370881e-06,
+      "loss": 0.1851,
+      "step": 29020
+    },
+    {
+      "epoch": 0.3821656050955414,
+      "grad_norm": 1.4650992155075073,
+      "learning_rate": 7.085389613018707e-06,
+      "loss": 0.2083,
+      "step": 29040
+    },
+    {
+      "epoch": 0.38242880454808653,
+      "grad_norm": 1.7759548425674438,
+      "learning_rate": 7.081515054300104e-06,
+      "loss": 0.1903,
+      "step": 29060
+    },
+    {
+      "epoch": 0.3826920040006317,
+      "grad_norm": 1.0641263723373413,
+      "learning_rate": 7.077638983030551e-06,
+      "loss": 0.1883,
+      "step": 29080
+    },
+    {
+      "epoch": 0.3829552034531768,
+      "grad_norm": 1.2633455991744995,
+      "learning_rate": 7.073761402026632e-06,
+      "loss": 0.2121,
+      "step": 29100
+    },
+    {
+      "epoch": 0.38321840290572196,
+      "grad_norm": 1.647993564605713,
+      "learning_rate": 7.069882314106022e-06,
+      "loss": 0.1805,
+      "step": 29120
+    },
+    {
+      "epoch": 0.3834816023582671,
+      "grad_norm": 1.3624417781829834,
+      "learning_rate": 7.066001722087496e-06,
+      "loss": 0.1794,
+      "step": 29140
+    },
+    {
+      "epoch": 0.3837448018108122,
+      "grad_norm": 1.7590969800949097,
+      "learning_rate": 7.062119628790918e-06,
+      "loss": 0.1876,
+      "step": 29160
+    },
+    {
+      "epoch": 0.3840080012633574,
+      "grad_norm": 1.805687427520752,
+      "learning_rate": 7.058236037037245e-06,
+      "loss": 0.2188,
+      "step": 29180
+    },
+    {
+      "epoch": 0.3842712007159025,
+      "grad_norm": 1.8855760097503662,
+      "learning_rate": 7.054350949648523e-06,
+      "loss": 0.1835,
+      "step": 29200
+    },
+    {
+      "epoch": 0.38453440016844764,
+      "grad_norm": 1.5083848237991333,
+      "learning_rate": 7.050464369447885e-06,
+      "loss": 0.1889,
+      "step": 29220
+    },
+    {
+      "epoch": 0.3847975996209928,
+      "grad_norm": 1.6822813749313354,
+      "learning_rate": 7.046576299259546e-06,
+      "loss": 0.2017,
+      "step": 29240
+    },
+    {
+      "epoch": 0.3850607990735379,
+      "grad_norm": 1.774574875831604,
+      "learning_rate": 7.042686741908808e-06,
+      "loss": 0.2009,
+      "step": 29260
+    },
+    {
+      "epoch": 0.38532399852608307,
+      "grad_norm": 1.5243043899536133,
+      "learning_rate": 7.038795700222054e-06,
+      "loss": 0.1943,
+      "step": 29280
+    },
+    {
+      "epoch": 0.3855871979786282,
+      "grad_norm": 1.5482676029205322,
+      "learning_rate": 7.034903177026738e-06,
+      "loss": 0.1954,
+      "step": 29300
+    },
+    {
+      "epoch": 0.38585039743117333,
+      "grad_norm": 1.3015425205230713,
+      "learning_rate": 7.031009175151398e-06,
+      "loss": 0.1986,
+      "step": 29320
+    },
+    {
+      "epoch": 0.3861135968837185,
+      "grad_norm": 1.404215693473816,
+      "learning_rate": 7.027113697425643e-06,
+      "loss": 0.1825,
+      "step": 29340
+    },
+    {
+      "epoch": 0.3863767963362636,
+      "grad_norm": 1.5641545057296753,
+      "learning_rate": 7.023216746680157e-06,
+      "loss": 0.1877,
+      "step": 29360
+    },
+    {
+      "epoch": 0.38663999578880875,
+      "grad_norm": 1.7373216152191162,
+      "learning_rate": 7.019318325746694e-06,
+      "loss": 0.1816,
+      "step": 29380
+    },
+    {
+      "epoch": 0.3869031952413539,
+      "grad_norm": 0.8709298968315125,
+      "learning_rate": 7.015418437458073e-06,
+      "loss": 0.1992,
+      "step": 29400
+    },
+    {
+      "epoch": 0.387166394693899,
+      "grad_norm": 1.5133869647979736,
+      "learning_rate": 7.011517084648183e-06,
+      "loss": 0.1814,
+      "step": 29420
+    },
+    {
+      "epoch": 0.3874295941464442,
+      "grad_norm": 5.625399589538574,
+      "learning_rate": 7.007614270151977e-06,
+      "loss": 0.2001,
+      "step": 29440
+    },
+    {
+      "epoch": 0.38769279359898934,
+      "grad_norm": 1.913761854171753,
+      "learning_rate": 7.0037099968054655e-06,
+      "loss": 0.1816,
+      "step": 29460
+    },
+    {
+      "epoch": 0.38795599305153444,
+      "grad_norm": 1.1597191095352173,
+      "learning_rate": 6.999804267445726e-06,
+      "loss": 0.1972,
+      "step": 29480
+    },
+    {
+      "epoch": 0.3882191925040796,
+      "grad_norm": 1.7580773830413818,
+      "learning_rate": 6.995897084910888e-06,
+      "loss": 0.1869,
+      "step": 29500
+    },
+    {
+      "epoch": 0.3882191925040796,
+      "eval_loss": 0.22335517406463623,
+      "eval_runtime": 3.5654,
+      "eval_samples_per_second": 4.488,
+      "eval_steps_per_second": 2.244,
+      "step": 29500
+    },
+    {
+      "epoch": 0.3884823919566247,
+      "grad_norm": 1.5400073528289795,
+      "learning_rate": 6.991988452040142e-06,
+      "loss": 0.197,
+      "step": 29520
+    },
+    {
+      "epoch": 0.38874559140916987,
+      "grad_norm": 1.5688652992248535,
+      "learning_rate": 6.988078371673726e-06,
+      "loss": 0.1974,
+      "step": 29540
+    },
+    {
+      "epoch": 0.389008790861715,
+      "grad_norm": 1.5705808401107788,
+      "learning_rate": 6.98416684665294e-06,
+      "loss": 0.1836,
+      "step": 29560
+    },
+    {
+      "epoch": 0.38927199031426013,
+      "grad_norm": 1.6469228267669678,
+      "learning_rate": 6.980253879820121e-06,
+      "loss": 0.1958,
+      "step": 29580
+    },
+    {
+      "epoch": 0.3895351897668053,
+      "grad_norm": 1.7418570518493652,
+      "learning_rate": 6.9763394740186655e-06,
+      "loss": 0.1989,
+      "step": 29600
+    },
+    {
+      "epoch": 0.38979838921935045,
+      "grad_norm": 1.4512022733688354,
+      "learning_rate": 6.9724236320930064e-06,
+      "loss": 0.1985,
+      "step": 29620
+    },
+    {
+      "epoch": 0.39006158867189555,
+      "grad_norm": 1.3726671934127808,
+      "learning_rate": 6.9685063568886265e-06,
+      "loss": 0.2023,
+      "step": 29640
+    },
+    {
+      "epoch": 0.3903247881244407,
+      "grad_norm": 1.917883038520813,
+      "learning_rate": 6.964587651252047e-06,
+      "loss": 0.1881,
+      "step": 29660
+    },
+    {
+      "epoch": 0.3905879875769858,
+      "grad_norm": 1.4018638134002686,
+      "learning_rate": 6.9606675180308305e-06,
+      "loss": 0.1899,
+      "step": 29680
+    },
+    {
+      "epoch": 0.390851187029531,
+      "grad_norm": 1.9795598983764648,
+      "learning_rate": 6.956745960073572e-06,
+      "loss": 0.2041,
+      "step": 29700
+    },
+    {
+      "epoch": 0.39111438648207614,
+      "grad_norm": 1.8003655672073364,
+      "learning_rate": 6.952822980229911e-06,
+      "loss": 0.2005,
+      "step": 29720
+    },
+    {
+      "epoch": 0.39137758593462124,
+      "grad_norm": 1.005361557006836,
+      "learning_rate": 6.94889858135051e-06,
+      "loss": 0.1938,
+      "step": 29740
+    },
+    {
+      "epoch": 0.3916407853871664,
+      "grad_norm": 1.0519945621490479,
+      "learning_rate": 6.9449727662870706e-06,
+      "loss": 0.1975,
+      "step": 29760
+    },
+    {
+      "epoch": 0.39190398483971156,
+      "grad_norm": 1.6894583702087402,
+      "learning_rate": 6.941045537892319e-06,
+      "loss": 0.1829,
+      "step": 29780
+    },
+    {
+      "epoch": 0.39216718429225667,
+      "grad_norm": 1.5923384428024292,
+      "learning_rate": 6.93711689902001e-06,
+      "loss": 0.182,
+      "step": 29800
+    },
+    {
+      "epoch": 0.3924303837448018,
+      "grad_norm": 1.4540765285491943,
+      "learning_rate": 6.933186852524924e-06,
+      "loss": 0.2029,
+      "step": 29820
+    },
+    {
+      "epoch": 0.39269358319734693,
+      "grad_norm": 1.4255832433700562,
+      "learning_rate": 6.929255401262861e-06,
+      "loss": 0.2061,
+      "step": 29840
+    },
+    {
+      "epoch": 0.3929567826498921,
+      "grad_norm": 1.524120569229126,
+      "learning_rate": 6.925322548090646e-06,
+      "loss": 0.1828,
+      "step": 29860
+    },
+    {
+      "epoch": 0.39321998210243725,
+      "grad_norm": 1.7158714532852173,
+      "learning_rate": 6.921388295866119e-06,
+      "loss": 0.1839,
+      "step": 29880
+    },
+    {
+      "epoch": 0.39348318155498235,
+      "grad_norm": 1.342188835144043,
+      "learning_rate": 6.917452647448143e-06,
+      "loss": 0.212,
+      "step": 29900
+    },
+    {
+      "epoch": 0.3937463810075275,
+      "grad_norm": 1.5715876817703247,
+      "learning_rate": 6.913515605696584e-06,
+      "loss": 0.1805,
+      "step": 29920
+    },
+    {
+      "epoch": 0.3940095804600726,
+      "grad_norm": 1.4209752082824707,
+      "learning_rate": 6.909577173472335e-06,
+      "loss": 0.2084,
+      "step": 29940
+    },
+    {
+      "epoch": 0.3942727799126178,
+      "grad_norm": 1.1319451332092285,
+      "learning_rate": 6.905637353637285e-06,
+      "loss": 0.2067,
+      "step": 29960
+    },
+    {
+      "epoch": 0.39453597936516294,
+      "grad_norm": 1.3922033309936523,
+      "learning_rate": 6.901696149054345e-06,
+      "loss": 0.1963,
+      "step": 29980
+    },
+    {
+      "epoch": 0.39479917881770804,
+      "grad_norm": 1.7566359043121338,
+      "learning_rate": 6.897753562587418e-06,
+      "loss": 0.192,
+      "step": 30000
+    },
+    {
+      "epoch": 0.39479917881770804,
+      "eval_loss": 0.2217421978712082,
+      "eval_runtime": 3.5511,
+      "eval_samples_per_second": 4.506,
+      "eval_steps_per_second": 2.253,
+      "step": 30000
+    },
+    {
+      "epoch": 0.3950623782702532,
+      "grad_norm": 1.0749309062957764,
+      "learning_rate": 6.893809597101424e-06,
+      "loss": 0.2121,
+      "step": 30020
+    },
+    {
+      "epoch": 0.39532557772279836,
+      "grad_norm": 1.5761021375656128,
+      "learning_rate": 6.889864255462279e-06,
+      "loss": 0.198,
+      "step": 30040
+    },
+    {
+      "epoch": 0.39558877717534346,
+      "grad_norm": 1.7993015050888062,
+      "learning_rate": 6.885917540536897e-06,
+      "loss": 0.1992,
+      "step": 30060
+    },
+    {
+      "epoch": 0.3958519766278886,
+      "grad_norm": 1.0883554220199585,
+      "learning_rate": 6.881969455193197e-06,
+      "loss": 0.2074,
+      "step": 30080
+    },
+    {
+      "epoch": 0.39611517608043373,
+      "grad_norm": 1.8873635530471802,
+      "learning_rate": 6.8780200023000855e-06,
+      "loss": 0.2274,
+      "step": 30100
+    },
+    {
+      "epoch": 0.3963783755329789,
+      "grad_norm": 1.6381438970565796,
+      "learning_rate": 6.8740691847274675e-06,
+      "loss": 0.1924,
+      "step": 30120
+    },
+    {
+      "epoch": 0.39664157498552405,
+      "grad_norm": 1.1293567419052124,
+      "learning_rate": 6.870117005346241e-06,
+      "loss": 0.1875,
+      "step": 30140
+    },
+    {
+      "epoch": 0.39690477443806915,
+      "grad_norm": 1.4629812240600586,
+      "learning_rate": 6.8661634670282885e-06,
+      "loss": 0.1946,
+      "step": 30160
+    },
+    {
+      "epoch": 0.3971679738906143,
+      "grad_norm": 1.6992765665054321,
+      "learning_rate": 6.862208572646485e-06,
+      "loss": 0.1871,
+      "step": 30180
+    },
+    {
+      "epoch": 0.39743117334315947,
+      "grad_norm": 1.6246271133422852,
+      "learning_rate": 6.858252325074686e-06,
+      "loss": 0.1866,
+      "step": 30200
+    },
+    {
+      "epoch": 0.3976943727957046,
+      "grad_norm": 1.338696002960205,
+      "learning_rate": 6.854294727187737e-06,
+      "loss": 0.2077,
+      "step": 30220
+    },
+    {
+      "epoch": 0.39795757224824974,
+      "grad_norm": 1.3836256265640259,
+      "learning_rate": 6.850335781861457e-06,
+      "loss": 0.2072,
+      "step": 30240
+    },
+    {
+      "epoch": 0.39822077170079484,
+      "grad_norm": 1.1888128519058228,
+      "learning_rate": 6.8463754919726525e-06,
+      "loss": 0.1935,
+      "step": 30260
+    },
+    {
+      "epoch": 0.39848397115334,
+      "grad_norm": 1.590481162071228,
+      "learning_rate": 6.842413860399097e-06,
+      "loss": 0.1787,
+      "step": 30280
+    },
+    {
+      "epoch": 0.39874717060588516,
+      "grad_norm": 1.1861306428909302,
+      "learning_rate": 6.838450890019548e-06,
+      "loss": 0.1832,
+      "step": 30300
+    },
+    {
+      "epoch": 0.39901037005843026,
+      "grad_norm": 1.6327004432678223,
+      "learning_rate": 6.834486583713731e-06,
+      "loss": 0.1764,
+      "step": 30320
+    },
+    {
+      "epoch": 0.3992735695109754,
+      "grad_norm": 1.3854563236236572,
+      "learning_rate": 6.830520944362344e-06,
+      "loss": 0.1875,
+      "step": 30340
+    },
+    {
+      "epoch": 0.3995367689635206,
+      "grad_norm": 1.7575511932373047,
+      "learning_rate": 6.826553974847054e-06,
+      "loss": 0.2036,
+      "step": 30360
+    },
+    {
+      "epoch": 0.3997999684160657,
+      "grad_norm": 1.7319692373275757,
+      "learning_rate": 6.822585678050492e-06,
+      "loss": 0.1929,
+      "step": 30380
+    },
+    {
+      "epoch": 0.40006316786861085,
+      "grad_norm": 1.7108440399169922,
+      "learning_rate": 6.8186160568562555e-06,
+      "loss": 0.1949,
+      "step": 30400
+    },
+    {
+      "epoch": 0.40032636732115595,
+      "grad_norm": 1.6456578969955444,
+      "learning_rate": 6.814645114148906e-06,
+      "loss": 0.191,
+      "step": 30420
+    },
+    {
+      "epoch": 0.4005895667737011,
+      "grad_norm": 1.419944167137146,
+      "learning_rate": 6.810672852813962e-06,
+      "loss": 0.1992,
+      "step": 30440
+    },
+    {
+      "epoch": 0.40085276622624627,
+      "grad_norm": 1.5517572164535522,
+      "learning_rate": 6.8066992757379e-06,
+      "loss": 0.1796,
+      "step": 30460
+    },
+    {
+      "epoch": 0.4011159656787914,
+      "grad_norm": 1.385994553565979,
+      "learning_rate": 6.8027243858081575e-06,
+      "loss": 0.1917,
+      "step": 30480
+    },
+    {
+      "epoch": 0.40137916513133653,
+      "grad_norm": 1.6804977655410767,
+      "learning_rate": 6.79874818591312e-06,
+      "loss": 0.1833,
+      "step": 30500
+    },
+    {
+      "epoch": 0.40137916513133653,
+      "eval_loss": 0.22047245502471924,
+      "eval_runtime": 3.6409,
+      "eval_samples_per_second": 4.395,
+      "eval_steps_per_second": 2.197,
+      "step": 30500
+    },
+    {
+      "epoch": 0.40164236458388164,
+      "grad_norm": 1.6051151752471924,
+      "learning_rate": 6.7947706789421305e-06,
+      "loss": 0.2117,
+      "step": 30520
+    },
+    {
+      "epoch": 0.4019055640364268,
+      "grad_norm": 1.302005648612976,
+      "learning_rate": 6.790791867785475e-06,
+      "loss": 0.1835,
+      "step": 30540
+    },
+    {
+      "epoch": 0.40216876348897196,
+      "grad_norm": 1.165794849395752,
+      "learning_rate": 6.7868117553343946e-06,
+      "loss": 0.2057,
+      "step": 30560
+    },
+    {
+      "epoch": 0.40243196294151706,
+      "grad_norm": 1.2935985326766968,
+      "learning_rate": 6.78283034448107e-06,
+      "loss": 0.1795,
+      "step": 30580
+    },
+    {
+      "epoch": 0.4026951623940622,
+      "grad_norm": 1.3392261266708374,
+      "learning_rate": 6.778847638118633e-06,
+      "loss": 0.2093,
+      "step": 30600
+    },
+    {
+      "epoch": 0.4029583618466074,
+      "grad_norm": 1.2944869995117188,
+      "learning_rate": 6.774863639141143e-06,
+      "loss": 0.194,
+      "step": 30620
+    },
+    {
+      "epoch": 0.4032215612991525,
+      "grad_norm": 1.8727186918258667,
+      "learning_rate": 6.770878350443615e-06,
+      "loss": 0.1901,
+      "step": 30640
+    },
+    {
+      "epoch": 0.40348476075169765,
+      "grad_norm": 1.6432219743728638,
+      "learning_rate": 6.766891774921989e-06,
+      "loss": 0.187,
+      "step": 30660
+    },
+    {
+      "epoch": 0.40374796020424275,
+      "grad_norm": 1.4779781103134155,
+      "learning_rate": 6.762903915473147e-06,
+      "loss": 0.2032,
+      "step": 30680
+    },
+    {
+      "epoch": 0.4040111596567879,
+      "grad_norm": 1.7180678844451904,
+      "learning_rate": 6.7589147749949006e-06,
+      "loss": 0.1859,
+      "step": 30700
+    },
+    {
+      "epoch": 0.40427435910933307,
+      "grad_norm": 1.3678371906280518,
+      "learning_rate": 6.754924356385993e-06,
+      "loss": 0.1922,
+      "step": 30720
+    },
+    {
+      "epoch": 0.4045375585618782,
+      "grad_norm": 1.4549665451049805,
+      "learning_rate": 6.750932662546097e-06,
+      "loss": 0.1883,
+      "step": 30740
+    },
+    {
+      "epoch": 0.40480075801442333,
+      "grad_norm": 1.8011565208435059,
+      "learning_rate": 6.746939696375814e-06,
+      "loss": 0.1973,
+      "step": 30760
+    },
+    {
+      "epoch": 0.4050639574669685,
+      "grad_norm": 1.7316049337387085,
+      "learning_rate": 6.742945460776663e-06,
+      "loss": 0.1848,
+      "step": 30780
+    },
+    {
+      "epoch": 0.4053271569195136,
+      "grad_norm": 1.832383155822754,
+      "learning_rate": 6.738949958651091e-06,
+      "loss": 0.1828,
+      "step": 30800
+    },
+    {
+      "epoch": 0.40559035637205876,
+      "grad_norm": 1.6263402700424194,
+      "learning_rate": 6.734953192902467e-06,
+      "loss": 0.1934,
+      "step": 30820
+    },
+    {
+      "epoch": 0.40585355582460386,
+      "grad_norm": 1.3486895561218262,
+      "learning_rate": 6.730955166435074e-06,
+      "loss": 0.1955,
+      "step": 30840
+    },
+    {
+      "epoch": 0.406116755277149,
+      "grad_norm": 1.5041359663009644,
+      "learning_rate": 6.726955882154112e-06,
+      "loss": 0.1957,
+      "step": 30860
+    },
+    {
+      "epoch": 0.4063799547296942,
+      "grad_norm": 1.2625662088394165,
+      "learning_rate": 6.722955342965694e-06,
+      "loss": 0.1952,
+      "step": 30880
+    },
+    {
+      "epoch": 0.4066431541822393,
+      "grad_norm": 1.2576124668121338,
+      "learning_rate": 6.7189535517768505e-06,
+      "loss": 0.1809,
+      "step": 30900
+    },
+    {
+      "epoch": 0.40690635363478445,
+      "grad_norm": 1.0093426704406738,
+      "learning_rate": 6.714950511495512e-06,
+      "loss": 0.1801,
+      "step": 30920
+    },
+    {
+      "epoch": 0.4071695530873296,
+      "grad_norm": 1.5277131795883179,
+      "learning_rate": 6.7109462250305295e-06,
+      "loss": 0.2103,
+      "step": 30940
+    },
+    {
+      "epoch": 0.4074327525398747,
+      "grad_norm": 1.4498957395553589,
+      "learning_rate": 6.706940695291648e-06,
+      "loss": 0.1993,
+      "step": 30960
+    },
+    {
+      "epoch": 0.40769595199241987,
+      "grad_norm": 1.5070521831512451,
+      "learning_rate": 6.702933925189521e-06,
+      "loss": 0.2022,
+      "step": 30980
+    },
+    {
+      "epoch": 0.407959151444965,
+      "grad_norm": 1.335569143295288,
+      "learning_rate": 6.698925917635701e-06,
+      "loss": 0.1869,
+      "step": 31000
+    },
+    {
+      "epoch": 0.407959151444965,
+      "eval_loss": 0.2227441668510437,
+      "eval_runtime": 3.6222,
+      "eval_samples_per_second": 4.417,
+      "eval_steps_per_second": 2.209,
+      "step": 31000
+    },
+    {
+      "epoch": 0.40822235089751013,
+      "grad_norm": 2.0715415477752686,
+      "learning_rate": 6.694916675542646e-06,
+      "loss": 0.1986,
+      "step": 31020
+    },
+    {
+      "epoch": 0.4084855503500553,
+      "grad_norm": 1.426714301109314,
+      "learning_rate": 6.690906201823704e-06,
+      "loss": 0.1974,
+      "step": 31040
+    },
+    {
+      "epoch": 0.4087487498026004,
+      "grad_norm": 2.2001657485961914,
+      "learning_rate": 6.68689449939312e-06,
+      "loss": 0.1999,
+      "step": 31060
+    },
+    {
+      "epoch": 0.40901194925514556,
+      "grad_norm": 1.8032435178756714,
+      "learning_rate": 6.682881571166031e-06,
+      "loss": 0.1843,
+      "step": 31080
+    },
+    {
+      "epoch": 0.40927514870769066,
+      "grad_norm": 1.5007072687149048,
+      "learning_rate": 6.678867420058468e-06,
+      "loss": 0.2044,
+      "step": 31100
+    },
+    {
+      "epoch": 0.4095383481602358,
+      "grad_norm": 1.8214564323425293,
+      "learning_rate": 6.674852048987351e-06,
+      "loss": 0.2008,
+      "step": 31120
+    },
+    {
+      "epoch": 0.409801547612781,
+      "grad_norm": 1.5820116996765137,
+      "learning_rate": 6.67083546087048e-06,
+      "loss": 0.1965,
+      "step": 31140
+    },
+    {
+      "epoch": 0.4100647470653261,
+      "grad_norm": 1.4120547771453857,
+      "learning_rate": 6.666817658626543e-06,
+      "loss": 0.1934,
+      "step": 31160
+    },
+    {
+      "epoch": 0.41032794651787124,
+      "grad_norm": 1.4382630586624146,
+      "learning_rate": 6.662798645175116e-06,
+      "loss": 0.1808,
+      "step": 31180
+    },
+    {
+      "epoch": 0.4105911459704164,
+      "grad_norm": 1.2393957376480103,
+      "learning_rate": 6.6587784234366445e-06,
+      "loss": 0.1592,
+      "step": 31200
+    },
+    {
+      "epoch": 0.4108543454229615,
+      "grad_norm": 1.5943717956542969,
+      "learning_rate": 6.6547569963324635e-06,
+      "loss": 0.1985,
+      "step": 31220
+    },
+    {
+      "epoch": 0.41111754487550667,
+      "grad_norm": 1.3115501403808594,
+      "learning_rate": 6.65073436678477e-06,
+      "loss": 0.197,
+      "step": 31240
+    },
+    {
+      "epoch": 0.4113807443280518,
+      "grad_norm": 1.4647728204727173,
+      "learning_rate": 6.64671053771665e-06,
+      "loss": 0.1916,
+      "step": 31260
+    },
+    {
+      "epoch": 0.41164394378059693,
+      "grad_norm": 1.67393159866333,
+      "learning_rate": 6.6426855120520475e-06,
+      "loss": 0.1934,
+      "step": 31280
+    },
+    {
+      "epoch": 0.4119071432331421,
+      "grad_norm": 1.3441698551177979,
+      "learning_rate": 6.6386592927157875e-06,
+      "loss": 0.1794,
+      "step": 31300
+    },
+    {
+      "epoch": 0.4121703426856872,
+      "grad_norm": 1.7594759464263916,
+      "learning_rate": 6.6346318826335554e-06,
+      "loss": 0.2013,
+      "step": 31320
+    },
+    {
+      "epoch": 0.41243354213823236,
+      "grad_norm": 1.4499236345291138,
+      "learning_rate": 6.630603284731903e-06,
+      "loss": 0.1818,
+      "step": 31340
+    },
+    {
+      "epoch": 0.4126967415907775,
+      "grad_norm": 2.4897689819335938,
+      "learning_rate": 6.626573501938244e-06,
+      "loss": 0.1914,
+      "step": 31360
+    },
+    {
+      "epoch": 0.4129599410433226,
+      "grad_norm": 1.7486263513565063,
+      "learning_rate": 6.622542537180859e-06,
+      "loss": 0.2046,
+      "step": 31380
+    },
+    {
+      "epoch": 0.4132231404958678,
+      "grad_norm": 1.9230555295944214,
+      "learning_rate": 6.6185103933888814e-06,
+      "loss": 0.1857,
+      "step": 31400
+    },
+    {
+      "epoch": 0.4134863399484129,
+      "grad_norm": 1.4280365705490112,
+      "learning_rate": 6.6144770734923024e-06,
+      "loss": 0.193,
+      "step": 31420
+    },
+    {
+      "epoch": 0.41374953940095804,
+      "grad_norm": 1.3191720247268677,
+      "learning_rate": 6.610442580421972e-06,
+      "loss": 0.1883,
+      "step": 31440
+    },
+    {
+      "epoch": 0.4140127388535032,
+      "grad_norm": 1.7624657154083252,
+      "learning_rate": 6.606406917109587e-06,
+      "loss": 0.1908,
+      "step": 31460
+    },
+    {
+      "epoch": 0.4142759383060483,
+      "grad_norm": 1.5302268266677856,
+      "learning_rate": 6.602370086487697e-06,
+      "loss": 0.1846,
+      "step": 31480
+    },
+    {
+      "epoch": 0.41453913775859347,
+      "grad_norm": 1.3704450130462646,
+      "learning_rate": 6.5983320914897e-06,
+      "loss": 0.1738,
+      "step": 31500
+    },
+    {
+      "epoch": 0.41453913775859347,
+      "eval_loss": 0.22191569209098816,
+      "eval_runtime": 3.6592,
+      "eval_samples_per_second": 4.373,
+      "eval_steps_per_second": 2.186,
+      "step": 31500
+    },
+    {
+      "epoch": 0.4148023372111386,
+      "grad_norm": 1.5241117477416992,
+      "learning_rate": 6.594292935049842e-06,
+      "loss": 0.1864,
+      "step": 31520
+    },
+    {
+      "epoch": 0.41506553666368373,
+      "grad_norm": 1.8416653871536255,
+      "learning_rate": 6.590252620103208e-06,
+      "loss": 0.1965,
+      "step": 31540
+    },
+    {
+      "epoch": 0.4153287361162289,
+      "grad_norm": 1.50454843044281,
+      "learning_rate": 6.586211149585733e-06,
+      "loss": 0.169,
+      "step": 31560
+    },
+    {
+      "epoch": 0.415591935568774,
+      "grad_norm": 1.6529409885406494,
+      "learning_rate": 6.58216852643418e-06,
+      "loss": 0.194,
+      "step": 31580
+    },
+    {
+      "epoch": 0.41585513502131916,
+      "grad_norm": 1.398409366607666,
+      "learning_rate": 6.5781247535861604e-06,
+      "loss": 0.2038,
+      "step": 31600
+    },
+    {
+      "epoch": 0.4161183344738643,
+      "grad_norm": 1.3833739757537842,
+      "learning_rate": 6.574079833980115e-06,
+      "loss": 0.1903,
+      "step": 31620
+    },
+    {
+      "epoch": 0.4163815339264094,
+      "grad_norm": 1.197349190711975,
+      "learning_rate": 6.570033770555321e-06,
+      "loss": 0.1924,
+      "step": 31640
+    },
+    {
+      "epoch": 0.4166447333789546,
+      "grad_norm": 2.130068063735962,
+      "learning_rate": 6.565986566251885e-06,
+      "loss": 0.192,
+      "step": 31660
+    },
+    {
+      "epoch": 0.4169079328314997,
+      "grad_norm": 1.5700234174728394,
+      "learning_rate": 6.5619382240107425e-06,
+      "loss": 0.1958,
+      "step": 31680
+    },
+    {
+      "epoch": 0.41717113228404484,
+      "grad_norm": 1.344245195388794,
+      "learning_rate": 6.557888746773655e-06,
+      "loss": 0.1907,
+      "step": 31700
+    },
+    {
+      "epoch": 0.41743433173659,
+      "grad_norm": 1.4489773511886597,
+      "learning_rate": 6.553838137483215e-06,
+      "loss": 0.1892,
+      "step": 31720
+    },
+    {
+      "epoch": 0.4176975311891351,
+      "grad_norm": 1.1369203329086304,
+      "learning_rate": 6.549786399082828e-06,
+      "loss": 0.1931,
+      "step": 31740
+    },
+    {
+      "epoch": 0.41796073064168027,
+      "grad_norm": 2.0791735649108887,
+      "learning_rate": 6.545733534516727e-06,
+      "loss": 0.201,
+      "step": 31760
+    },
+    {
+      "epoch": 0.4182239300942254,
+      "grad_norm": 1.6038734912872314,
+      "learning_rate": 6.541679546729961e-06,
+      "loss": 0.1842,
+      "step": 31780
+    },
+    {
+      "epoch": 0.41848712954677053,
+      "grad_norm": 1.347807765007019,
+      "learning_rate": 6.537624438668395e-06,
+      "loss": 0.1706,
+      "step": 31800
+    },
+    {
+      "epoch": 0.4187503289993157,
+      "grad_norm": 1.2694398164749146,
+      "learning_rate": 6.53356821327871e-06,
+      "loss": 0.1836,
+      "step": 31820
+    },
+    {
+      "epoch": 0.4190135284518608,
+      "grad_norm": 1.2187561988830566,
+      "learning_rate": 6.529510873508394e-06,
+      "loss": 0.2018,
+      "step": 31840
+    },
+    {
+      "epoch": 0.41927672790440595,
+      "grad_norm": 2.1708993911743164,
+      "learning_rate": 6.525452422305751e-06,
+      "loss": 0.184,
+      "step": 31860
+    },
+    {
+      "epoch": 0.4195399273569511,
+      "grad_norm": 1.5449150800704956,
+      "learning_rate": 6.521392862619889e-06,
+      "loss": 0.1929,
+      "step": 31880
+    },
+    {
+      "epoch": 0.4198031268094962,
+      "grad_norm": 1.3529999256134033,
+      "learning_rate": 6.5173321974007195e-06,
+      "loss": 0.1794,
+      "step": 31900
+    },
+    {
+      "epoch": 0.4200663262620414,
+      "grad_norm": 1.527464747428894,
+      "learning_rate": 6.513270429598966e-06,
+      "loss": 0.1948,
+      "step": 31920
+    },
+    {
+      "epoch": 0.42032952571458654,
+      "grad_norm": 0.8968510031700134,
+      "learning_rate": 6.5092075621661425e-06,
+      "loss": 0.1872,
+      "step": 31940
+    },
+    {
+      "epoch": 0.42059272516713164,
+      "grad_norm": 1.6669769287109375,
+      "learning_rate": 6.5051435980545664e-06,
+      "loss": 0.1894,
+      "step": 31960
+    },
+    {
+      "epoch": 0.4208559246196768,
+      "grad_norm": 1.2000232934951782,
+      "learning_rate": 6.501078540217356e-06,
+      "loss": 0.1786,
+      "step": 31980
+    },
+    {
+      "epoch": 0.4211191240722219,
+      "grad_norm": 1.6172252893447876,
+      "learning_rate": 6.497012391608418e-06,
+      "loss": 0.1776,
+      "step": 32000
+    },
+    {
+      "epoch": 0.4211191240722219,
+      "eval_loss": 0.22333350777626038,
+      "eval_runtime": 4.4928,
+      "eval_samples_per_second": 3.561,
+      "eval_steps_per_second": 1.781,
+      "step": 32000
+    },
+    {
+      "epoch": 0.42138232352476707,
+      "grad_norm": 1.6331032514572144,
+      "learning_rate": 6.4929451551824565e-06,
+      "loss": 0.2044,
+      "step": 32020
+    },
+    {
+      "epoch": 0.4216455229773122,
+      "grad_norm": 1.2594256401062012,
+      "learning_rate": 6.488876833894961e-06,
+      "loss": 0.1839,
+      "step": 32040
+    },
+    {
+      "epoch": 0.42190872242985733,
+      "grad_norm": 2.0694472789764404,
+      "learning_rate": 6.484807430702216e-06,
+      "loss": 0.1734,
+      "step": 32060
+    },
+    {
+      "epoch": 0.4221719218824025,
+      "grad_norm": 1.490417242050171,
+      "learning_rate": 6.480736948561286e-06,
+      "loss": 0.1878,
+      "step": 32080
+    },
+    {
+      "epoch": 0.42243512133494765,
+      "grad_norm": 1.1691862344741821,
+      "learning_rate": 6.476665390430026e-06,
+      "loss": 0.1813,
+      "step": 32100
+    },
+    {
+      "epoch": 0.42269832078749275,
+      "grad_norm": 1.431425929069519,
+      "learning_rate": 6.472592759267062e-06,
+      "loss": 0.1998,
+      "step": 32120
+    },
+    {
+      "epoch": 0.4229615202400379,
+      "grad_norm": 1.4938523769378662,
+      "learning_rate": 6.468519058031815e-06,
+      "loss": 0.1751,
+      "step": 32140
+    },
+    {
+      "epoch": 0.423224719692583,
+      "grad_norm": 1.3803164958953857,
+      "learning_rate": 6.464444289684467e-06,
+      "loss": 0.194,
+      "step": 32160
+    },
+    {
+      "epoch": 0.4234879191451282,
+      "grad_norm": 1.3079752922058105,
+      "learning_rate": 6.460368457185992e-06,
+      "loss": 0.2026,
+      "step": 32180
+    },
+    {
+      "epoch": 0.42375111859767334,
+      "grad_norm": 2.530695676803589,
+      "learning_rate": 6.456291563498122e-06,
+      "loss": 0.1958,
+      "step": 32200
+    },
+    {
+      "epoch": 0.42401431805021844,
+      "grad_norm": 1.8456814289093018,
+      "learning_rate": 6.452213611583371e-06,
+      "loss": 0.196,
+      "step": 32220
+    },
+    {
+      "epoch": 0.4242775175027636,
+      "grad_norm": 1.3038290739059448,
+      "learning_rate": 6.448134604405016e-06,
+      "loss": 0.1883,
+      "step": 32240
+    },
+    {
+      "epoch": 0.4245407169553087,
+      "grad_norm": 1.1577657461166382,
+      "learning_rate": 6.4440545449271065e-06,
+      "loss": 0.1777,
+      "step": 32260
+    },
+    {
+      "epoch": 0.42480391640785387,
+      "grad_norm": 1.3067224025726318,
+      "learning_rate": 6.4399734361144485e-06,
+      "loss": 0.1946,
+      "step": 32280
+    },
+    {
+      "epoch": 0.425067115860399,
+      "grad_norm": 1.4227482080459595,
+      "learning_rate": 6.4358912809326165e-06,
+      "loss": 0.1859,
+      "step": 32300
+    },
+    {
+      "epoch": 0.42533031531294413,
+      "grad_norm": 1.142237663269043,
+      "learning_rate": 6.431808082347945e-06,
+      "loss": 0.1973,
+      "step": 32320
+    },
+    {
+      "epoch": 0.4255935147654893,
+      "grad_norm": 1.6217082738876343,
+      "learning_rate": 6.427723843327527e-06,
+      "loss": 0.1893,
+      "step": 32340
+    },
+    {
+      "epoch": 0.42585671421803445,
+      "grad_norm": 1.406288504600525,
+      "learning_rate": 6.423638566839207e-06,
+      "loss": 0.1963,
+      "step": 32360
+    },
+    {
+      "epoch": 0.42611991367057955,
+      "grad_norm": 1.256482720375061,
+      "learning_rate": 6.419552255851589e-06,
+      "loss": 0.2104,
+      "step": 32380
+    },
+    {
+      "epoch": 0.4263831131231247,
+      "grad_norm": 1.6554683446884155,
+      "learning_rate": 6.415464913334023e-06,
+      "loss": 0.1874,
+      "step": 32400
+    },
+    {
+      "epoch": 0.4266463125756698,
+      "grad_norm": 1.2006601095199585,
+      "learning_rate": 6.411376542256615e-06,
+      "loss": 0.1947,
+      "step": 32420
+    },
+    {
+      "epoch": 0.426909512028215,
+      "grad_norm": 1.0604151487350464,
+      "learning_rate": 6.407287145590215e-06,
+      "loss": 0.1807,
+      "step": 32440
+    },
+    {
+      "epoch": 0.42717271148076014,
+      "grad_norm": 1.672268271446228,
+      "learning_rate": 6.403196726306418e-06,
+      "loss": 0.1922,
+      "step": 32460
+    },
+    {
+      "epoch": 0.42743591093330524,
+      "grad_norm": 1.6486583948135376,
+      "learning_rate": 6.399105287377562e-06,
+      "loss": 0.2112,
+      "step": 32480
+    },
+    {
+      "epoch": 0.4276991103858504,
+      "grad_norm": 3.2525100708007812,
+      "learning_rate": 6.395012831776727e-06,
+      "loss": 0.1796,
+      "step": 32500
+    },
+    {
+      "epoch": 0.4276991103858504,
+      "eval_loss": 0.21887081861495972,
+      "eval_runtime": 4.3687,
+      "eval_samples_per_second": 3.662,
+      "eval_steps_per_second": 1.831,
+      "step": 32500
+    },
+    {
+      "epoch": 0.42796230983839556,
+      "grad_norm": 1.3204389810562134,
+      "learning_rate": 6.390919362477732e-06,
+      "loss": 0.1875,
+      "step": 32520
+    },
+    {
+      "epoch": 0.42822550929094066,
+      "grad_norm": 1.499685287475586,
+      "learning_rate": 6.386824882455129e-06,
+      "loss": 0.198,
+      "step": 32540
+    },
+    {
+      "epoch": 0.4284887087434858,
+      "grad_norm": 1.837647557258606,
+      "learning_rate": 6.382729394684212e-06,
+      "loss": 0.184,
+      "step": 32560
+    },
+    {
+      "epoch": 0.42875190819603093,
+      "grad_norm": 1.3775413036346436,
+      "learning_rate": 6.378632902140999e-06,
+      "loss": 0.1746,
+      "step": 32580
+    },
+    {
+      "epoch": 0.4290151076485761,
+      "grad_norm": 1.9418611526489258,
+      "learning_rate": 6.3745354078022435e-06,
+      "loss": 0.1993,
+      "step": 32600
+    },
+    {
+      "epoch": 0.42927830710112125,
+      "grad_norm": 1.670531153678894,
+      "learning_rate": 6.370436914645425e-06,
+      "loss": 0.1851,
+      "step": 32620
+    },
+    {
+      "epoch": 0.42954150655366635,
+      "grad_norm": 1.7601070404052734,
+      "learning_rate": 6.366337425648749e-06,
+      "loss": 0.1864,
+      "step": 32640
+    },
+    {
+      "epoch": 0.4298047060062115,
+      "grad_norm": 1.6713740825653076,
+      "learning_rate": 6.3622369437911446e-06,
+      "loss": 0.1798,
+      "step": 32660
+    },
+    {
+      "epoch": 0.43006790545875667,
+      "grad_norm": 1.4265562295913696,
+      "learning_rate": 6.358135472052263e-06,
+      "loss": 0.186,
+      "step": 32680
+    },
+    {
+      "epoch": 0.4303311049113018,
+      "grad_norm": 1.3544732332229614,
+      "learning_rate": 6.354033013412473e-06,
+      "loss": 0.1881,
+      "step": 32700
+    },
+    {
+      "epoch": 0.43059430436384694,
+      "grad_norm": 1.4449738264083862,
+      "learning_rate": 6.349929570852862e-06,
+      "loss": 0.1913,
+      "step": 32720
+    },
+    {
+      "epoch": 0.43085750381639204,
+      "grad_norm": 1.3270151615142822,
+      "learning_rate": 6.345825147355232e-06,
+      "loss": 0.1712,
+      "step": 32740
+    },
+    {
+      "epoch": 0.4311207032689372,
+      "grad_norm": 1.4760222434997559,
+      "learning_rate": 6.3417197459021e-06,
+      "loss": 0.1776,
+      "step": 32760
+    },
+    {
+      "epoch": 0.43138390272148236,
+      "grad_norm": 1.06393301486969,
+      "learning_rate": 6.337613369476689e-06,
+      "loss": 0.2042,
+      "step": 32780
+    },
+    {
+      "epoch": 0.43164710217402746,
+      "grad_norm": 1.1887221336364746,
+      "learning_rate": 6.333506021062935e-06,
+      "loss": 0.1906,
+      "step": 32800
+    },
+    {
+      "epoch": 0.4319103016265726,
+      "grad_norm": 1.3478950262069702,
+      "learning_rate": 6.329397703645475e-06,
+      "loss": 0.1859,
+      "step": 32820
+    },
+    {
+      "epoch": 0.43217350107911773,
+      "grad_norm": 1.1891019344329834,
+      "learning_rate": 6.325288420209658e-06,
+      "loss": 0.2049,
+      "step": 32840
+    },
+    {
+      "epoch": 0.4324367005316629,
+      "grad_norm": 1.142300009727478,
+      "learning_rate": 6.321178173741527e-06,
+      "loss": 0.1802,
+      "step": 32860
+    },
+    {
+      "epoch": 0.43269989998420805,
+      "grad_norm": 1.2861328125,
+      "learning_rate": 6.317066967227829e-06,
+      "loss": 0.1955,
+      "step": 32880
+    },
+    {
+      "epoch": 0.43296309943675315,
+      "grad_norm": 1.7539379596710205,
+      "learning_rate": 6.312954803656008e-06,
+      "loss": 0.1831,
+      "step": 32900
+    },
+    {
+      "epoch": 0.4332262988892983,
+      "grad_norm": 1.3429447412490845,
+      "learning_rate": 6.308841686014204e-06,
+      "loss": 0.1918,
+      "step": 32920
+    },
+    {
+      "epoch": 0.43348949834184347,
+      "grad_norm": 1.9349175691604614,
+      "learning_rate": 6.3047276172912486e-06,
+      "loss": 0.2031,
+      "step": 32940
+    },
+    {
+      "epoch": 0.4337526977943886,
+      "grad_norm": 1.7988762855529785,
+      "learning_rate": 6.300612600476666e-06,
+      "loss": 0.1876,
+      "step": 32960
+    },
+    {
+      "epoch": 0.43401589724693374,
+      "grad_norm": 1.3601737022399902,
+      "learning_rate": 6.296496638560669e-06,
+      "loss": 0.1864,
+      "step": 32980
+    },
+    {
+      "epoch": 0.43427909669947884,
+      "grad_norm": 1.915542483329773,
+      "learning_rate": 6.292379734534157e-06,
+      "loss": 0.1725,
+      "step": 33000
+    },
+    {
+      "epoch": 0.43427909669947884,
+      "eval_loss": 0.22344526648521423,
+      "eval_runtime": 4.2302,
+      "eval_samples_per_second": 3.782,
+      "eval_steps_per_second": 1.891,
+      "step": 33000
+    },
+    {
+      "epoch": 0.434542296152024,
+      "grad_norm": 1.559260606765747,
+      "learning_rate": 6.288261891388713e-06,
+      "loss": 0.178,
+      "step": 33020
+    },
+    {
+      "epoch": 0.43480549560456916,
+      "grad_norm": 1.2293295860290527,
+      "learning_rate": 6.284143112116605e-06,
+      "loss": 0.1888,
+      "step": 33040
+    },
+    {
+      "epoch": 0.43506869505711426,
+      "grad_norm": 1.7175774574279785,
+      "learning_rate": 6.280023399710777e-06,
+      "loss": 0.1791,
+      "step": 33060
+    },
+    {
+      "epoch": 0.4353318945096594,
+      "grad_norm": 1.7191330194473267,
+      "learning_rate": 6.275902757164856e-06,
+      "loss": 0.2009,
+      "step": 33080
+    },
+    {
+      "epoch": 0.4355950939622046,
+      "grad_norm": 1.3217905759811401,
+      "learning_rate": 6.271781187473143e-06,
+      "loss": 0.2,
+      "step": 33100
+    },
+    {
+      "epoch": 0.4358582934147497,
+      "grad_norm": 1.1840640306472778,
+      "learning_rate": 6.267658693630612e-06,
+      "loss": 0.1908,
+      "step": 33120
+    },
+    {
+      "epoch": 0.43612149286729485,
+      "grad_norm": 1.6475924253463745,
+      "learning_rate": 6.263535278632906e-06,
+      "loss": 0.1902,
+      "step": 33140
+    },
+    {
+      "epoch": 0.43638469231983995,
+      "grad_norm": 1.3678683042526245,
+      "learning_rate": 6.259410945476343e-06,
+      "loss": 0.1949,
+      "step": 33160
+    },
+    {
+      "epoch": 0.4366478917723851,
+      "grad_norm": 1.2859419584274292,
+      "learning_rate": 6.255285697157905e-06,
+      "loss": 0.1798,
+      "step": 33180
+    },
+    {
+      "epoch": 0.43691109122493027,
+      "grad_norm": 1.9422056674957275,
+      "learning_rate": 6.251159536675239e-06,
+      "loss": 0.2056,
+      "step": 33200
+    },
+    {
+      "epoch": 0.4371742906774754,
+      "grad_norm": 1.5574455261230469,
+      "learning_rate": 6.247032467026657e-06,
+      "loss": 0.1894,
+      "step": 33220
+    },
+    {
+      "epoch": 0.43743749013002053,
+      "grad_norm": 1.3285733461380005,
+      "learning_rate": 6.242904491211127e-06,
+      "loss": 0.1899,
+      "step": 33240
+    },
+    {
+      "epoch": 0.4377006895825657,
+      "grad_norm": 1.7193233966827393,
+      "learning_rate": 6.238775612228281e-06,
+      "loss": 0.2013,
+      "step": 33260
+    },
+    {
+      "epoch": 0.4379638890351108,
+      "grad_norm": 1.9625941514968872,
+      "learning_rate": 6.234645833078401e-06,
+      "loss": 0.202,
+      "step": 33280
+    },
+    {
+      "epoch": 0.43822708848765596,
+      "grad_norm": 1.6575123071670532,
+      "learning_rate": 6.230515156762432e-06,
+      "loss": 0.1923,
+      "step": 33300
+    },
+    {
+      "epoch": 0.43849028794020106,
+      "grad_norm": 1.5500762462615967,
+      "learning_rate": 6.2263835862819635e-06,
+      "loss": 0.1902,
+      "step": 33320
+    },
+    {
+      "epoch": 0.4387534873927462,
+      "grad_norm": 1.6159378290176392,
+      "learning_rate": 6.2222511246392344e-06,
+      "loss": 0.1696,
+      "step": 33340
+    },
+    {
+      "epoch": 0.4390166868452914,
+      "grad_norm": 1.3241300582885742,
+      "learning_rate": 6.218117774837137e-06,
+      "loss": 0.1835,
+      "step": 33360
+    },
+    {
+      "epoch": 0.4392798862978365,
+      "grad_norm": 1.531108021736145,
+      "learning_rate": 6.213983539879205e-06,
+      "loss": 0.1994,
+      "step": 33380
+    },
+    {
+      "epoch": 0.43954308575038165,
+      "grad_norm": 1.5252314805984497,
+      "learning_rate": 6.209848422769616e-06,
+      "loss": 0.1906,
+      "step": 33400
+    },
+    {
+      "epoch": 0.4398062852029268,
+      "grad_norm": 1.585817575454712,
+      "learning_rate": 6.205712426513187e-06,
+      "loss": 0.1773,
+      "step": 33420
+    },
+    {
+      "epoch": 0.4400694846554719,
+      "grad_norm": 1.8318239450454712,
+      "learning_rate": 6.201575554115378e-06,
+      "loss": 0.1771,
+      "step": 33440
+    },
+    {
+      "epoch": 0.44033268410801707,
+      "grad_norm": 1.3142973184585571,
+      "learning_rate": 6.197437808582281e-06,
+      "loss": 0.1749,
+      "step": 33460
+    },
+    {
+      "epoch": 0.4405958835605622,
+      "grad_norm": 1.1685500144958496,
+      "learning_rate": 6.193299192920626e-06,
+      "loss": 0.2258,
+      "step": 33480
+    },
+    {
+      "epoch": 0.44085908301310733,
+      "grad_norm": 1.3037041425704956,
+      "learning_rate": 6.1891597101377756e-06,
+      "loss": 0.1918,
+      "step": 33500
+    },
+    {
+      "epoch": 0.44085908301310733,
+      "eval_loss": 0.21774965524673462,
+      "eval_runtime": 4.1514,
+      "eval_samples_per_second": 3.854,
+      "eval_steps_per_second": 1.927,
+      "step": 33500
+    },
+    {
+      "epoch": 0.4411222824656525,
+      "grad_norm": 1.5754035711288452,
+      "learning_rate": 6.185019363241716e-06,
+      "loss": 0.2048,
+      "step": 33520
+    },
+    {
+      "epoch": 0.4413854819181976,
+      "grad_norm": 1.0297467708587646,
+      "learning_rate": 6.1808781552410715e-06,
+      "loss": 0.1772,
+      "step": 33540
+    },
+    {
+      "epoch": 0.44164868137074276,
+      "grad_norm": 1.6313148736953735,
+      "learning_rate": 6.176736089145082e-06,
+      "loss": 0.1878,
+      "step": 33560
+    },
+    {
+      "epoch": 0.44191188082328786,
+      "grad_norm": 1.2680325508117676,
+      "learning_rate": 6.17259316796362e-06,
+      "loss": 0.1955,
+      "step": 33580
+    },
+    {
+      "epoch": 0.442175080275833,
+      "grad_norm": 1.2366173267364502,
+      "learning_rate": 6.168449394707172e-06,
+      "loss": 0.1815,
+      "step": 33600
+    },
+    {
+      "epoch": 0.4424382797283782,
+      "grad_norm": 1.8294126987457275,
+      "learning_rate": 6.164304772386849e-06,
+      "loss": 0.1794,
+      "step": 33620
+    },
+    {
+      "epoch": 0.4427014791809233,
+      "grad_norm": 1.5196397304534912,
+      "learning_rate": 6.160159304014374e-06,
+      "loss": 0.1877,
+      "step": 33640
+    },
+    {
+      "epoch": 0.44296467863346844,
+      "grad_norm": 1.5132142305374146,
+      "learning_rate": 6.156012992602089e-06,
+      "loss": 0.1891,
+      "step": 33660
+    },
+    {
+      "epoch": 0.4432278780860136,
+      "grad_norm": 1.2361456155776978,
+      "learning_rate": 6.151865841162948e-06,
+      "loss": 0.1809,
+      "step": 33680
+    },
+    {
+      "epoch": 0.4434910775385587,
+      "grad_norm": 1.90122652053833,
+      "learning_rate": 6.147717852710513e-06,
+      "loss": 0.1799,
+      "step": 33700
+    },
+    {
+      "epoch": 0.44375427699110387,
+      "grad_norm": 2.3359997272491455,
+      "learning_rate": 6.143569030258958e-06,
+      "loss": 0.1968,
+      "step": 33720
+    },
+    {
+      "epoch": 0.444017476443649,
+      "grad_norm": 1.537970781326294,
+      "learning_rate": 6.139419376823057e-06,
+      "loss": 0.1944,
+      "step": 33740
+    },
+    {
+      "epoch": 0.44428067589619413,
+      "grad_norm": 1.889626145362854,
+      "learning_rate": 6.135268895418196e-06,
+      "loss": 0.1759,
+      "step": 33760
+    },
+    {
+      "epoch": 0.4445438753487393,
+      "grad_norm": 1.4113043546676636,
+      "learning_rate": 6.131117589060354e-06,
+      "loss": 0.1786,
+      "step": 33780
+    },
+    {
+      "epoch": 0.4448070748012844,
+      "grad_norm": 1.5016658306121826,
+      "learning_rate": 6.126965460766119e-06,
+      "loss": 0.1695,
+      "step": 33800
+    },
+    {
+      "epoch": 0.44507027425382956,
+      "grad_norm": 2.474191665649414,
+      "learning_rate": 6.122812513552664e-06,
+      "loss": 0.1708,
+      "step": 33820
+    },
+    {
+      "epoch": 0.4453334737063747,
+      "grad_norm": 1.5038944482803345,
+      "learning_rate": 6.118658750437771e-06,
+      "loss": 0.1846,
+      "step": 33840
+    },
+    {
+      "epoch": 0.4455966731589198,
+      "grad_norm": 1.3580965995788574,
+      "learning_rate": 6.114504174439804e-06,
+      "loss": 0.1809,
+      "step": 33860
+    },
+    {
+      "epoch": 0.445859872611465,
+      "grad_norm": 1.428220272064209,
+      "learning_rate": 6.110348788577722e-06,
+      "loss": 0.1982,
+      "step": 33880
+    },
+    {
+      "epoch": 0.4461230720640101,
+      "grad_norm": 2.1817941665649414,
+      "learning_rate": 6.106192595871072e-06,
+      "loss": 0.194,
+      "step": 33900
+    },
+    {
+      "epoch": 0.44638627151655524,
+      "grad_norm": 1.5743972063064575,
+      "learning_rate": 6.102035599339991e-06,
+      "loss": 0.2123,
+      "step": 33920
+    },
+    {
+      "epoch": 0.4466494709691004,
+      "grad_norm": 1.8085507154464722,
+      "learning_rate": 6.09787780200519e-06,
+      "loss": 0.1865,
+      "step": 33940
+    },
+    {
+      "epoch": 0.4469126704216455,
+      "grad_norm": 1.4177377223968506,
+      "learning_rate": 6.093719206887974e-06,
+      "loss": 0.1839,
+      "step": 33960
+    },
+    {
+      "epoch": 0.44717586987419067,
+      "grad_norm": 1.382252812385559,
+      "learning_rate": 6.089559817010219e-06,
+      "loss": 0.1891,
+      "step": 33980
+    },
+    {
+      "epoch": 0.4474390693267358,
+      "grad_norm": 1.6128599643707275,
+      "learning_rate": 6.085399635394384e-06,
+      "loss": 0.178,
+      "step": 34000
+    },
+    {
+      "epoch": 0.4474390693267358,
+      "eval_loss": 0.22218140959739685,
+      "eval_runtime": 4.3168,
+      "eval_samples_per_second": 3.706,
+      "eval_steps_per_second": 1.853,
+      "step": 34000
+    },
+    {
+      "epoch": 0.44770226877928093,
+      "grad_norm": 1.7131582498550415,
+      "learning_rate": 6.081238665063501e-06,
+      "loss": 0.1899,
+      "step": 34020
+    },
+    {
+      "epoch": 0.4479654682318261,
+      "grad_norm": 1.3275742530822754,
+      "learning_rate": 6.077076909041173e-06,
+      "loss": 0.1989,
+      "step": 34040
+    },
+    {
+      "epoch": 0.4482286676843712,
+      "grad_norm": 1.2196422815322876,
+      "learning_rate": 6.072914370351576e-06,
+      "loss": 0.1992,
+      "step": 34060
+    },
+    {
+      "epoch": 0.44849186713691636,
+      "grad_norm": 1.8276389837265015,
+      "learning_rate": 6.068751052019457e-06,
+      "loss": 0.1793,
+      "step": 34080
+    },
+    {
+      "epoch": 0.4487550665894615,
+      "grad_norm": 1.564966082572937,
+      "learning_rate": 6.064586957070126e-06,
+      "loss": 0.193,
+      "step": 34100
+    },
+    {
+      "epoch": 0.4490182660420066,
+      "grad_norm": 1.819327473640442,
+      "learning_rate": 6.060422088529459e-06,
+      "loss": 0.1904,
+      "step": 34120
+    },
+    {
+      "epoch": 0.4492814654945518,
+      "grad_norm": 1.7045491933822632,
+      "learning_rate": 6.056256449423895e-06,
+      "loss": 0.1862,
+      "step": 34140
+    },
+    {
+      "epoch": 0.4495446649470969,
+      "grad_norm": 1.4048614501953125,
+      "learning_rate": 6.052090042780431e-06,
+      "loss": 0.1896,
+      "step": 34160
+    },
+    {
+      "epoch": 0.44980786439964204,
+      "grad_norm": 1.4375282526016235,
+      "learning_rate": 6.0479228716266195e-06,
+      "loss": 0.1907,
+      "step": 34180
+    },
+    {
+      "epoch": 0.4500710638521872,
+      "grad_norm": 1.0854368209838867,
+      "learning_rate": 6.043754938990576e-06,
+      "loss": 0.1827,
+      "step": 34200
+    },
+    {
+      "epoch": 0.4503342633047323,
+      "grad_norm": 1.7280949354171753,
+      "learning_rate": 6.039586247900962e-06,
+      "loss": 0.1886,
+      "step": 34220
+    },
+    {
+      "epoch": 0.45059746275727747,
+      "grad_norm": 1.4985793828964233,
+      "learning_rate": 6.035416801386994e-06,
+      "loss": 0.176,
+      "step": 34240
+    },
+    {
+      "epoch": 0.4508606622098226,
+      "grad_norm": 1.417004942893982,
+      "learning_rate": 6.0312466024784375e-06,
+      "loss": 0.1902,
+      "step": 34260
+    },
+    {
+      "epoch": 0.45112386166236773,
+      "grad_norm": 1.3839576244354248,
+      "learning_rate": 6.027075654205602e-06,
+      "loss": 0.1959,
+      "step": 34280
+    },
+    {
+      "epoch": 0.4513870611149129,
+      "grad_norm": 1.5660208463668823,
+      "learning_rate": 6.022903959599344e-06,
+      "loss": 0.1965,
+      "step": 34300
+    },
+    {
+      "epoch": 0.451650260567458,
+      "grad_norm": 2.1012418270111084,
+      "learning_rate": 6.018731521691059e-06,
+      "loss": 0.1856,
+      "step": 34320
+    },
+    {
+      "epoch": 0.45191346002000315,
+      "grad_norm": 1.5790983438491821,
+      "learning_rate": 6.01455834351269e-06,
+      "loss": 0.1805,
+      "step": 34340
+    },
+    {
+      "epoch": 0.4521766594725483,
+      "grad_norm": 1.4756594896316528,
+      "learning_rate": 6.010384428096709e-06,
+      "loss": 0.1936,
+      "step": 34360
+    },
+    {
+      "epoch": 0.4524398589250934,
+      "grad_norm": 1.5107929706573486,
+      "learning_rate": 6.006209778476131e-06,
+      "loss": 0.1739,
+      "step": 34380
+    },
+    {
+      "epoch": 0.4527030583776386,
+      "grad_norm": 1.3859968185424805,
+      "learning_rate": 6.002034397684497e-06,
+      "loss": 0.2058,
+      "step": 34400
+    },
+    {
+      "epoch": 0.45296625783018374,
+      "grad_norm": 1.819901466369629,
+      "learning_rate": 5.997858288755889e-06,
+      "loss": 0.1916,
+      "step": 34420
+    },
+    {
+      "epoch": 0.45322945728272884,
+      "grad_norm": 1.4252551794052124,
+      "learning_rate": 5.9936814547249075e-06,
+      "loss": 0.188,
+      "step": 34440
+    },
+    {
+      "epoch": 0.453492656735274,
+      "grad_norm": 1.7155195474624634,
+      "learning_rate": 5.98950389862669e-06,
+      "loss": 0.1818,
+      "step": 34460
+    },
+    {
+      "epoch": 0.4537558561878191,
+      "grad_norm": 1.6819077730178833,
+      "learning_rate": 5.98532562349689e-06,
+      "loss": 0.2053,
+      "step": 34480
+    },
+    {
+      "epoch": 0.45401905564036427,
+      "grad_norm": 1.592030644416809,
+      "learning_rate": 5.981146632371689e-06,
+      "loss": 0.1714,
+      "step": 34500
+    },
+    {
+      "epoch": 0.45401905564036427,
+      "eval_loss": 0.21568642556667328,
+      "eval_runtime": 4.3061,
+      "eval_samples_per_second": 3.716,
+      "eval_steps_per_second": 1.858,
+      "step": 34500
+    },
+    {
+      "epoch": 0.4542822550929094,
+      "grad_norm": 1.5019720792770386,
+      "learning_rate": 5.976966928287786e-06,
+      "loss": 0.1858,
+      "step": 34520
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 1.3714592456817627,
+      "learning_rate": 5.972786514282402e-06,
+      "loss": 0.1812,
+      "step": 34540
+    },
+    {
+      "epoch": 0.4548086539979997,
+      "grad_norm": 1.8025953769683838,
+      "learning_rate": 5.968605393393268e-06,
+      "loss": 0.2176,
+      "step": 34560
+    },
+    {
+      "epoch": 0.45507185345054485,
+      "grad_norm": 1.7548881769180298,
+      "learning_rate": 5.964423568658634e-06,
+      "loss": 0.2058,
+      "step": 34580
+    },
+    {
+      "epoch": 0.45533505290308995,
+      "grad_norm": 1.346596121788025,
+      "learning_rate": 5.960241043117256e-06,
+      "loss": 0.1982,
+      "step": 34600
+    },
+    {
+      "epoch": 0.4555982523556351,
+      "grad_norm": 1.905904769897461,
+      "learning_rate": 5.956057819808405e-06,
+      "loss": 0.1933,
+      "step": 34620
+    },
+    {
+      "epoch": 0.4558614518081802,
+      "grad_norm": 1.4718680381774902,
+      "learning_rate": 5.951873901771858e-06,
+      "loss": 0.186,
+      "step": 34640
+    },
+    {
+      "epoch": 0.4561246512607254,
+      "grad_norm": 1.517948865890503,
+      "learning_rate": 5.947689292047892e-06,
+      "loss": 0.1818,
+      "step": 34660
+    },
+    {
+      "epoch": 0.45638785071327054,
+      "grad_norm": 1.2172093391418457,
+      "learning_rate": 5.943503993677289e-06,
+      "loss": 0.1907,
+      "step": 34680
+    },
+    {
+      "epoch": 0.45665105016581564,
+      "grad_norm": 1.1601641178131104,
+      "learning_rate": 5.939318009701336e-06,
+      "loss": 0.1876,
+      "step": 34700
+    },
+    {
+      "epoch": 0.4569142496183608,
+      "grad_norm": 1.6679763793945312,
+      "learning_rate": 5.935131343161813e-06,
+      "loss": 0.1832,
+      "step": 34720
+    },
+    {
+      "epoch": 0.4571774490709059,
+      "grad_norm": 1.3841214179992676,
+      "learning_rate": 5.9309439971009965e-06,
+      "loss": 0.1789,
+      "step": 34740
+    },
+    {
+      "epoch": 0.45744064852345107,
+      "grad_norm": 1.5580097436904907,
+      "learning_rate": 5.92675597456166e-06,
+      "loss": 0.1906,
+      "step": 34760
+    },
+    {
+      "epoch": 0.4577038479759962,
+      "grad_norm": 1.3432923555374146,
+      "learning_rate": 5.922567278587066e-06,
+      "loss": 0.1865,
+      "step": 34780
+    },
+    {
+      "epoch": 0.45796704742854133,
+      "grad_norm": 1.8977059125900269,
+      "learning_rate": 5.918377912220964e-06,
+      "loss": 0.1806,
+      "step": 34800
+    },
+    {
+      "epoch": 0.4582302468810865,
+      "grad_norm": 1.603629469871521,
+      "learning_rate": 5.914187878507596e-06,
+      "loss": 0.1967,
+      "step": 34820
+    },
+    {
+      "epoch": 0.45849344633363165,
+      "grad_norm": 1.6049772500991821,
+      "learning_rate": 5.9099971804916845e-06,
+      "loss": 0.1871,
+      "step": 34840
+    },
+    {
+      "epoch": 0.45875664578617675,
+      "grad_norm": 1.4688942432403564,
+      "learning_rate": 5.905805821218438e-06,
+      "loss": 0.1883,
+      "step": 34860
+    },
+    {
+      "epoch": 0.4590198452387219,
+      "grad_norm": 1.4824858903884888,
+      "learning_rate": 5.901613803733545e-06,
+      "loss": 0.1994,
+      "step": 34880
+    },
+    {
+      "epoch": 0.459283044691267,
+      "grad_norm": 1.5024317502975464,
+      "learning_rate": 5.897421131083169e-06,
+      "loss": 0.1707,
+      "step": 34900
+    },
+    {
+      "epoch": 0.4595462441438122,
+      "grad_norm": 2.00276780128479,
+      "learning_rate": 5.893227806313953e-06,
+      "loss": 0.1773,
+      "step": 34920
+    },
+    {
+      "epoch": 0.45980944359635734,
+      "grad_norm": 1.6530702114105225,
+      "learning_rate": 5.88903383247301e-06,
+      "loss": 0.1945,
+      "step": 34940
+    },
+    {
+      "epoch": 0.46007264304890244,
+      "grad_norm": 1.7898263931274414,
+      "learning_rate": 5.884839212607934e-06,
+      "loss": 0.1855,
+      "step": 34960
+    },
+    {
+      "epoch": 0.4603358425014476,
+      "grad_norm": 1.458601713180542,
+      "learning_rate": 5.880643949766776e-06,
+      "loss": 0.1875,
+      "step": 34980
+    },
+    {
+      "epoch": 0.46059904195399276,
+      "grad_norm": 1.6567147970199585,
+      "learning_rate": 5.8764480469980625e-06,
+      "loss": 0.1962,
+      "step": 35000
+    },
+    {
+      "epoch": 0.46059904195399276,
+      "eval_loss": 0.21850907802581787,
+      "eval_runtime": 3.6776,
+      "eval_samples_per_second": 4.351,
+      "eval_steps_per_second": 2.175,
+      "step": 35000
+    },
+    {
+      "epoch": 0.46086224140653786,
+      "grad_norm": 1.5212217569351196,
+      "learning_rate": 5.87225150735078e-06,
+      "loss": 0.1906,
+      "step": 35020
+    },
+    {
+      "epoch": 0.461125440859083,
+      "grad_norm": 1.4192168712615967,
+      "learning_rate": 5.8680543338743845e-06,
+      "loss": 0.1871,
+      "step": 35040
+    },
+    {
+      "epoch": 0.46138864031162813,
+      "grad_norm": 1.7703577280044556,
+      "learning_rate": 5.8638565296187856e-06,
+      "loss": 0.1961,
+      "step": 35060
+    },
+    {
+      "epoch": 0.4616518397641733,
+      "grad_norm": 1.4745707511901855,
+      "learning_rate": 5.859658097634356e-06,
+      "loss": 0.1958,
+      "step": 35080
+    },
+    {
+      "epoch": 0.46191503921671845,
+      "grad_norm": 1.4054754972457886,
+      "learning_rate": 5.85545904097192e-06,
+      "loss": 0.1748,
+      "step": 35100
+    },
+    {
+      "epoch": 0.46217823866926355,
+      "grad_norm": 1.2827421426773071,
+      "learning_rate": 5.851259362682762e-06,
+      "loss": 0.1924,
+      "step": 35120
+    },
+    {
+      "epoch": 0.4624414381218087,
+      "grad_norm": 1.8071829080581665,
+      "learning_rate": 5.84705906581861e-06,
+      "loss": 0.2087,
+      "step": 35140
+    },
+    {
+      "epoch": 0.46270463757435387,
+      "grad_norm": 1.6840096712112427,
+      "learning_rate": 5.842858153431652e-06,
+      "loss": 0.1907,
+      "step": 35160
+    },
+    {
+      "epoch": 0.462967837026899,
+      "grad_norm": 1.7727038860321045,
+      "learning_rate": 5.838656628574513e-06,
+      "loss": 0.1953,
+      "step": 35180
+    },
+    {
+      "epoch": 0.46323103647944414,
+      "grad_norm": 1.4764364957809448,
+      "learning_rate": 5.834454494300268e-06,
+      "loss": 0.1826,
+      "step": 35200
+    },
+    {
+      "epoch": 0.46349423593198924,
+      "grad_norm": 1.4131611585617065,
+      "learning_rate": 5.8302517536624345e-06,
+      "loss": 0.2049,
+      "step": 35220
+    },
+    {
+      "epoch": 0.4637574353845344,
+      "grad_norm": 1.5941535234451294,
+      "learning_rate": 5.826048409714972e-06,
+      "loss": 0.2087,
+      "step": 35240
+    },
+    {
+      "epoch": 0.46402063483707956,
+      "grad_norm": 1.7059826850891113,
+      "learning_rate": 5.821844465512276e-06,
+      "loss": 0.1585,
+      "step": 35260
+    },
+    {
+      "epoch": 0.46428383428962466,
+      "grad_norm": 1.16872239112854,
+      "learning_rate": 5.817639924109179e-06,
+      "loss": 0.1996,
+      "step": 35280
+    },
+    {
+      "epoch": 0.4645470337421698,
+      "grad_norm": 1.5232117176055908,
+      "learning_rate": 5.8134347885609455e-06,
+      "loss": 0.201,
+      "step": 35300
+    },
+    {
+      "epoch": 0.46481023319471493,
+      "grad_norm": 1.1008819341659546,
+      "learning_rate": 5.809229061923276e-06,
+      "loss": 0.2082,
+      "step": 35320
+    },
+    {
+      "epoch": 0.4650734326472601,
+      "grad_norm": 2.071932077407837,
+      "learning_rate": 5.805022747252298e-06,
+      "loss": 0.2095,
+      "step": 35340
+    },
+    {
+      "epoch": 0.46533663209980525,
+      "grad_norm": 1.6625778675079346,
+      "learning_rate": 5.800815847604564e-06,
+      "loss": 0.1914,
+      "step": 35360
+    },
+    {
+      "epoch": 0.46559983155235035,
+      "grad_norm": 1.1498914957046509,
+      "learning_rate": 5.7966083660370575e-06,
+      "loss": 0.1927,
+      "step": 35380
+    },
+    {
+      "epoch": 0.4658630310048955,
+      "grad_norm": 1.8067699670791626,
+      "learning_rate": 5.792400305607179e-06,
+      "loss": 0.178,
+      "step": 35400
+    },
+    {
+      "epoch": 0.46612623045744067,
+      "grad_norm": 1.473732829093933,
+      "learning_rate": 5.7881916693727556e-06,
+      "loss": 0.1939,
+      "step": 35420
+    },
+    {
+      "epoch": 0.4663894299099858,
+      "grad_norm": 1.3022547960281372,
+      "learning_rate": 5.783982460392023e-06,
+      "loss": 0.1945,
+      "step": 35440
+    },
+    {
+      "epoch": 0.46665262936253094,
+      "grad_norm": 1.317526936531067,
+      "learning_rate": 5.779772681723647e-06,
+      "loss": 0.189,
+      "step": 35460
+    },
+    {
+      "epoch": 0.46691582881507604,
+      "grad_norm": 1.5430481433868408,
+      "learning_rate": 5.775562336426692e-06,
+      "loss": 0.1851,
+      "step": 35480
+    },
+    {
+      "epoch": 0.4671790282676212,
+      "grad_norm": 1.3153587579727173,
+      "learning_rate": 5.77135142756065e-06,
+      "loss": 0.2022,
+      "step": 35500
+    },
+    {
+      "epoch": 0.4671790282676212,
+      "eval_loss": 0.21716155111789703,
+      "eval_runtime": 3.6736,
+      "eval_samples_per_second": 4.355,
+      "eval_steps_per_second": 2.178,
+      "step": 35500
+    },
+    {
+      "epoch": 0.46744222772016636,
+      "grad_norm": 1.7921133041381836,
+      "learning_rate": 5.767139958185406e-06,
+      "loss": 0.1758,
+      "step": 35520
+    },
+    {
+      "epoch": 0.46770542717271146,
+      "grad_norm": 1.6272165775299072,
+      "learning_rate": 5.762927931361267e-06,
+      "loss": 0.1838,
+      "step": 35540
+    },
+    {
+      "epoch": 0.4679686266252566,
+      "grad_norm": 1.1926414966583252,
+      "learning_rate": 5.758715350148936e-06,
+      "loss": 0.1861,
+      "step": 35560
+    },
+    {
+      "epoch": 0.4682318260778018,
+      "grad_norm": 1.1950007677078247,
+      "learning_rate": 5.754502217609522e-06,
+      "loss": 0.2101,
+      "step": 35580
+    },
+    {
+      "epoch": 0.4684950255303469,
+      "grad_norm": 2.089634656906128,
+      "learning_rate": 5.750288536804536e-06,
+      "loss": 0.1885,
+      "step": 35600
+    },
+    {
+      "epoch": 0.46875822498289205,
+      "grad_norm": 1.344348430633545,
+      "learning_rate": 5.7460743107958824e-06,
+      "loss": 0.1987,
+      "step": 35620
+    },
+    {
+      "epoch": 0.46902142443543715,
+      "grad_norm": 1.7029154300689697,
+      "learning_rate": 5.741859542645867e-06,
+      "loss": 0.1826,
+      "step": 35640
+    },
+    {
+      "epoch": 0.4692846238879823,
+      "grad_norm": 1.737906575202942,
+      "learning_rate": 5.737644235417189e-06,
+      "loss": 0.1851,
+      "step": 35660
+    },
+    {
+      "epoch": 0.46954782334052747,
+      "grad_norm": 1.8846346139907837,
+      "learning_rate": 5.733428392172935e-06,
+      "loss": 0.1798,
+      "step": 35680
+    },
+    {
+      "epoch": 0.4698110227930726,
+      "grad_norm": 1.7714552879333496,
+      "learning_rate": 5.729212015976586e-06,
+      "loss": 0.1802,
+      "step": 35700
+    },
+    {
+      "epoch": 0.47007422224561773,
+      "grad_norm": 1.1663082838058472,
+      "learning_rate": 5.724995109892006e-06,
+      "loss": 0.1952,
+      "step": 35720
+    },
+    {
+      "epoch": 0.4703374216981629,
+      "grad_norm": 1.5752499103546143,
+      "learning_rate": 5.720777676983451e-06,
+      "loss": 0.1712,
+      "step": 35740
+    },
+    {
+      "epoch": 0.470600621150708,
+      "grad_norm": 1.5831170082092285,
+      "learning_rate": 5.7165597203155495e-06,
+      "loss": 0.1902,
+      "step": 35760
+    },
+    {
+      "epoch": 0.47086382060325316,
+      "grad_norm": 2.1365966796875,
+      "learning_rate": 5.712341242953321e-06,
+      "loss": 0.2024,
+      "step": 35780
+    },
+    {
+      "epoch": 0.47112702005579826,
+      "grad_norm": 1.2058302164077759,
+      "learning_rate": 5.708122247962154e-06,
+      "loss": 0.1749,
+      "step": 35800
+    },
+    {
+      "epoch": 0.4713902195083434,
+      "grad_norm": 1.2352806329727173,
+      "learning_rate": 5.703902738407823e-06,
+      "loss": 0.1835,
+      "step": 35820
+    },
+    {
+      "epoch": 0.4716534189608886,
+      "grad_norm": 1.397844672203064,
+      "learning_rate": 5.699682717356466e-06,
+      "loss": 0.1898,
+      "step": 35840
+    },
+    {
+      "epoch": 0.4719166184134337,
+      "grad_norm": 1.3629508018493652,
+      "learning_rate": 5.695462187874602e-06,
+      "loss": 0.1953,
+      "step": 35860
+    },
+    {
+      "epoch": 0.47217981786597885,
+      "grad_norm": 1.4949841499328613,
+      "learning_rate": 5.691241153029114e-06,
+      "loss": 0.1806,
+      "step": 35880
+    },
+    {
+      "epoch": 0.47244301731852395,
+      "grad_norm": 1.5086630582809448,
+      "learning_rate": 5.687019615887251e-06,
+      "loss": 0.1819,
+      "step": 35900
+    },
+    {
+      "epoch": 0.4727062167710691,
+      "grad_norm": 1.4073050022125244,
+      "learning_rate": 5.682797579516635e-06,
+      "loss": 0.1916,
+      "step": 35920
+    },
+    {
+      "epoch": 0.47296941622361427,
+      "grad_norm": 1.6387346982955933,
+      "learning_rate": 5.6785750469852395e-06,
+      "loss": 0.1717,
+      "step": 35940
+    },
+    {
+      "epoch": 0.4732326156761594,
+      "grad_norm": 1.8164538145065308,
+      "learning_rate": 5.6743520213614075e-06,
+      "loss": 0.1878,
+      "step": 35960
+    },
+    {
+      "epoch": 0.47349581512870453,
+      "grad_norm": 1.3350290060043335,
+      "learning_rate": 5.670128505713837e-06,
+      "loss": 0.2027,
+      "step": 35980
+    },
+    {
+      "epoch": 0.4737590145812497,
+      "grad_norm": 1.4673668146133423,
+      "learning_rate": 5.6659045031115815e-06,
+      "loss": 0.1836,
+      "step": 36000
+    },
+    {
+      "epoch": 0.4737590145812497,
+      "eval_loss": 0.22151857614517212,
+      "eval_runtime": 4.4473,
+      "eval_samples_per_second": 3.598,
+      "eval_steps_per_second": 1.799,
+      "step": 36000
+    },
+    {
+      "epoch": 0.4740222140337948,
+      "grad_norm": 1.5888128280639648,
+      "learning_rate": 5.66168001662405e-06,
+      "loss": 0.1947,
+      "step": 36020
+    },
+    {
+      "epoch": 0.47428541348633996,
+      "grad_norm": 1.9369794130325317,
+      "learning_rate": 5.657455049321001e-06,
+      "loss": 0.1931,
+      "step": 36040
+    },
+    {
+      "epoch": 0.47454861293888506,
+      "grad_norm": 0.9932168126106262,
+      "learning_rate": 5.653229604272544e-06,
+      "loss": 0.2128,
+      "step": 36060
+    },
+    {
+      "epoch": 0.4748118123914302,
+      "grad_norm": 1.5537850856781006,
+      "learning_rate": 5.649003684549135e-06,
+      "loss": 0.206,
+      "step": 36080
+    },
+    {
+      "epoch": 0.4750750118439754,
+      "grad_norm": 1.6342934370040894,
+      "learning_rate": 5.644777293221575e-06,
+      "loss": 0.186,
+      "step": 36100
+    },
+    {
+      "epoch": 0.4753382112965205,
+      "grad_norm": 1.542791724205017,
+      "learning_rate": 5.640550433361012e-06,
+      "loss": 0.1846,
+      "step": 36120
+    },
+    {
+      "epoch": 0.47560141074906565,
+      "grad_norm": 2.083831310272217,
+      "learning_rate": 5.636323108038924e-06,
+      "loss": 0.18,
+      "step": 36140
+    },
+    {
+      "epoch": 0.4758646102016108,
+      "grad_norm": 1.416603922843933,
+      "learning_rate": 5.632095320327137e-06,
+      "loss": 0.1717,
+      "step": 36160
+    },
+    {
+      "epoch": 0.4761278096541559,
+      "grad_norm": 1.9717650413513184,
+      "learning_rate": 5.627867073297808e-06,
+      "loss": 0.1791,
+      "step": 36180
+    },
+    {
+      "epoch": 0.47639100910670107,
+      "grad_norm": 1.6320618391036987,
+      "learning_rate": 5.623638370023432e-06,
+      "loss": 0.1791,
+      "step": 36200
+    },
+    {
+      "epoch": 0.4766542085592462,
+      "grad_norm": 1.5891484022140503,
+      "learning_rate": 5.61940921357683e-06,
+      "loss": 0.2114,
+      "step": 36220
+    },
+    {
+      "epoch": 0.47691740801179133,
+      "grad_norm": 1.5472825765609741,
+      "learning_rate": 5.615179607031157e-06,
+      "loss": 0.178,
+      "step": 36240
+    },
+    {
+      "epoch": 0.4771806074643365,
+      "grad_norm": 1.7292464971542358,
+      "learning_rate": 5.610949553459891e-06,
+      "loss": 0.1831,
+      "step": 36260
+    },
+    {
+      "epoch": 0.4774438069168816,
+      "grad_norm": 3.4386112689971924,
+      "learning_rate": 5.60671905593684e-06,
+      "loss": 0.198,
+      "step": 36280
+    },
+    {
+      "epoch": 0.47770700636942676,
+      "grad_norm": 1.9190590381622314,
+      "learning_rate": 5.602488117536131e-06,
+      "loss": 0.1883,
+      "step": 36300
+    },
+    {
+      "epoch": 0.4779702058219719,
+      "grad_norm": 1.8157165050506592,
+      "learning_rate": 5.5982567413322105e-06,
+      "loss": 0.173,
+      "step": 36320
+    },
+    {
+      "epoch": 0.478233405274517,
+      "grad_norm": 1.6940088272094727,
+      "learning_rate": 5.594024930399846e-06,
+      "loss": 0.1751,
+      "step": 36340
+    },
+    {
+      "epoch": 0.4784966047270622,
+      "grad_norm": 1.7773345708847046,
+      "learning_rate": 5.5897926878141195e-06,
+      "loss": 0.1893,
+      "step": 36360
+    },
+    {
+      "epoch": 0.4787598041796073,
+      "grad_norm": 1.4546605348587036,
+      "learning_rate": 5.585560016650427e-06,
+      "loss": 0.188,
+      "step": 36380
+    },
+    {
+      "epoch": 0.47902300363215244,
+      "grad_norm": 0.9864956736564636,
+      "learning_rate": 5.581326919984472e-06,
+      "loss": 0.1941,
+      "step": 36400
+    },
+    {
+      "epoch": 0.4792862030846976,
+      "grad_norm": 1.8064433336257935,
+      "learning_rate": 5.5770934008922765e-06,
+      "loss": 0.194,
+      "step": 36420
+    },
+    {
+      "epoch": 0.4795494025372427,
+      "grad_norm": 0.9595035910606384,
+      "learning_rate": 5.572859462450161e-06,
+      "loss": 0.2089,
+      "step": 36440
+    },
+    {
+      "epoch": 0.47981260198978787,
+      "grad_norm": 1.7231264114379883,
+      "learning_rate": 5.5686251077347545e-06,
+      "loss": 0.179,
+      "step": 36460
+    },
+    {
+      "epoch": 0.480075801442333,
+      "grad_norm": 1.5161991119384766,
+      "learning_rate": 5.564390339822985e-06,
+      "loss": 0.2044,
+      "step": 36480
+    },
+    {
+      "epoch": 0.48033900089487813,
+      "grad_norm": 1.5309852361679077,
+      "learning_rate": 5.560155161792086e-06,
+      "loss": 0.1948,
+      "step": 36500
+    },
+    {
+      "epoch": 0.48033900089487813,
+      "eval_loss": 0.21812458336353302,
+      "eval_runtime": 4.248,
+      "eval_samples_per_second": 3.767,
+      "eval_steps_per_second": 1.883,
+      "step": 36500
+    },
+    {
+      "epoch": 0.4806022003474233,
+      "grad_norm": 1.4897534847259521,
+      "learning_rate": 5.555919576719585e-06,
+      "loss": 0.224,
+      "step": 36520
+    },
+    {
+      "epoch": 0.4808653997999684,
+      "grad_norm": 1.2553812265396118,
+      "learning_rate": 5.551683587683308e-06,
+      "loss": 0.1901,
+      "step": 36540
+    },
+    {
+      "epoch": 0.48112859925251356,
+      "grad_norm": 1.506169319152832,
+      "learning_rate": 5.547447197761372e-06,
+      "loss": 0.1785,
+      "step": 36560
+    },
+    {
+      "epoch": 0.4813917987050587,
+      "grad_norm": 1.5440646409988403,
+      "learning_rate": 5.5432104100321856e-06,
+      "loss": 0.187,
+      "step": 36580
+    },
+    {
+      "epoch": 0.4816549981576038,
+      "grad_norm": 1.5174026489257812,
+      "learning_rate": 5.538973227574449e-06,
+      "loss": 0.1926,
+      "step": 36600
+    },
+    {
+      "epoch": 0.481918197610149,
+      "grad_norm": 1.3052458763122559,
+      "learning_rate": 5.534735653467147e-06,
+      "loss": 0.1843,
+      "step": 36620
+    },
+    {
+      "epoch": 0.4821813970626941,
+      "grad_norm": 1.4376171827316284,
+      "learning_rate": 5.5304976907895504e-06,
+      "loss": 0.1947,
+      "step": 36640
+    },
+    {
+      "epoch": 0.48244459651523924,
+      "grad_norm": 1.5614553689956665,
+      "learning_rate": 5.52625934262121e-06,
+      "loss": 0.1661,
+      "step": 36660
+    },
+    {
+      "epoch": 0.4827077959677844,
+      "grad_norm": 1.596636176109314,
+      "learning_rate": 5.522020612041957e-06,
+      "loss": 0.1916,
+      "step": 36680
+    },
+    {
+      "epoch": 0.4829709954203295,
+      "grad_norm": 1.6016077995300293,
+      "learning_rate": 5.517781502131905e-06,
+      "loss": 0.1857,
+      "step": 36700
+    },
+    {
+      "epoch": 0.48323419487287467,
+      "grad_norm": 1.642420768737793,
+      "learning_rate": 5.5135420159714395e-06,
+      "loss": 0.1979,
+      "step": 36720
+    },
+    {
+      "epoch": 0.4834973943254198,
+      "grad_norm": 1.4463794231414795,
+      "learning_rate": 5.509302156641217e-06,
+      "loss": 0.1985,
+      "step": 36740
+    },
+    {
+      "epoch": 0.48376059377796493,
+      "grad_norm": 1.5472724437713623,
+      "learning_rate": 5.50506192722217e-06,
+      "loss": 0.1848,
+      "step": 36760
+    },
+    {
+      "epoch": 0.4840237932305101,
+      "grad_norm": 1.8126918077468872,
+      "learning_rate": 5.500821330795497e-06,
+      "loss": 0.1986,
+      "step": 36780
+    },
+    {
+      "epoch": 0.4842869926830552,
+      "grad_norm": 1.4678752422332764,
+      "learning_rate": 5.496580370442666e-06,
+      "loss": 0.1826,
+      "step": 36800
+    },
+    {
+      "epoch": 0.48455019213560035,
+      "grad_norm": 1.277424693107605,
+      "learning_rate": 5.492339049245406e-06,
+      "loss": 0.18,
+      "step": 36820
+    },
+    {
+      "epoch": 0.4848133915881455,
+      "grad_norm": 1.6442404985427856,
+      "learning_rate": 5.4880973702857106e-06,
+      "loss": 0.2012,
+      "step": 36840
+    },
+    {
+      "epoch": 0.4850765910406906,
+      "grad_norm": 1.712299108505249,
+      "learning_rate": 5.483855336645833e-06,
+      "loss": 0.1781,
+      "step": 36860
+    },
+    {
+      "epoch": 0.4853397904932358,
+      "grad_norm": 1.4734786748886108,
+      "learning_rate": 5.479612951408279e-06,
+      "loss": 0.1787,
+      "step": 36880
+    },
+    {
+      "epoch": 0.48560298994578094,
+      "grad_norm": 1.354980230331421,
+      "learning_rate": 5.475370217655822e-06,
+      "loss": 0.186,
+      "step": 36900
+    },
+    {
+      "epoch": 0.48586618939832604,
+      "grad_norm": 1.4729458093643188,
+      "learning_rate": 5.471127138471478e-06,
+      "loss": 0.1809,
+      "step": 36920
+    },
+    {
+      "epoch": 0.4861293888508712,
+      "grad_norm": 1.5554184913635254,
+      "learning_rate": 5.466883716938516e-06,
+      "loss": 0.1671,
+      "step": 36940
+    },
+    {
+      "epoch": 0.4863925883034163,
+      "grad_norm": 1.5091872215270996,
+      "learning_rate": 5.4626399561404555e-06,
+      "loss": 0.1816,
+      "step": 36960
+    },
+    {
+      "epoch": 0.48665578775596147,
+      "grad_norm": 1.4918692111968994,
+      "learning_rate": 5.458395859161063e-06,
+      "loss": 0.1693,
+      "step": 36980
+    },
+    {
+      "epoch": 0.4869189872085066,
+      "grad_norm": 1.5322606563568115,
+      "learning_rate": 5.454151429084348e-06,
+      "loss": 0.1878,
+      "step": 37000
+    },
+    {
+      "epoch": 0.4869189872085066,
+      "eval_loss": 0.21230493485927582,
+      "eval_runtime": 4.1888,
+      "eval_samples_per_second": 3.82,
+      "eval_steps_per_second": 1.91,
+      "step": 37000
+    },
+    {
+      "epoch": 0.48718218666105173,
+      "grad_norm": 1.6552225351333618,
+      "learning_rate": 5.449906668994561e-06,
+      "loss": 0.1917,
+      "step": 37020
+    },
+    {
+      "epoch": 0.4874453861135969,
+      "grad_norm": 1.30204176902771,
+      "learning_rate": 5.4456615819761975e-06,
+      "loss": 0.2048,
+      "step": 37040
+    },
+    {
+      "epoch": 0.48770858556614205,
+      "grad_norm": 1.808292269706726,
+      "learning_rate": 5.441416171113981e-06,
+      "loss": 0.1915,
+      "step": 37060
+    },
+    {
+      "epoch": 0.48797178501868715,
+      "grad_norm": 1.8942376375198364,
+      "learning_rate": 5.437170439492883e-06,
+      "loss": 0.1771,
+      "step": 37080
+    },
+    {
+      "epoch": 0.4882349844712323,
+      "grad_norm": 1.858835220336914,
+      "learning_rate": 5.432924390198095e-06,
+      "loss": 0.196,
+      "step": 37100
+    },
+    {
+      "epoch": 0.4884981839237774,
+      "grad_norm": 1.5447791814804077,
+      "learning_rate": 5.428678026315049e-06,
+      "loss": 0.1808,
+      "step": 37120
+    },
+    {
+      "epoch": 0.4887613833763226,
+      "grad_norm": 1.20880925655365,
+      "learning_rate": 5.424431350929399e-06,
+      "loss": 0.1937,
+      "step": 37140
+    },
+    {
+      "epoch": 0.48902458282886774,
+      "grad_norm": 1.209153413772583,
+      "learning_rate": 5.420184367127034e-06,
+      "loss": 0.1764,
+      "step": 37160
+    },
+    {
+      "epoch": 0.48928778228141284,
+      "grad_norm": 1.6946725845336914,
+      "learning_rate": 5.415937077994055e-06,
+      "loss": 0.1797,
+      "step": 37180
+    },
+    {
+      "epoch": 0.489550981733958,
+      "grad_norm": 1.4642714262008667,
+      "learning_rate": 5.411689486616795e-06,
+      "loss": 0.1934,
+      "step": 37200
+    },
+    {
+      "epoch": 0.4898141811865031,
+      "grad_norm": 1.1186001300811768,
+      "learning_rate": 5.407441596081804e-06,
+      "loss": 0.1984,
+      "step": 37220
+    },
+    {
+      "epoch": 0.49007738063904827,
+      "grad_norm": 1.6922529935836792,
+      "learning_rate": 5.4031934094758465e-06,
+      "loss": 0.1939,
+      "step": 37240
+    },
+    {
+      "epoch": 0.4903405800915934,
+      "grad_norm": 1.3709731101989746,
+      "learning_rate": 5.398944929885906e-06,
+      "loss": 0.1917,
+      "step": 37260
+    },
+    {
+      "epoch": 0.49060377954413853,
+      "grad_norm": 1.49043607711792,
+      "learning_rate": 5.394696160399174e-06,
+      "loss": 0.1874,
+      "step": 37280
+    },
+    {
+      "epoch": 0.4908669789966837,
+      "grad_norm": 1.2977590560913086,
+      "learning_rate": 5.390447104103057e-06,
+      "loss": 0.1954,
+      "step": 37300
+    },
+    {
+      "epoch": 0.49113017844922885,
+      "grad_norm": 1.283549427986145,
+      "learning_rate": 5.38619776408517e-06,
+      "loss": 0.1822,
+      "step": 37320
+    },
+    {
+      "epoch": 0.49139337790177395,
+      "grad_norm": 1.5116095542907715,
+      "learning_rate": 5.381948143433332e-06,
+      "loss": 0.1726,
+      "step": 37340
+    },
+    {
+      "epoch": 0.4916565773543191,
+      "grad_norm": 1.5191696882247925,
+      "learning_rate": 5.377698245235566e-06,
+      "loss": 0.1946,
+      "step": 37360
+    },
+    {
+      "epoch": 0.4919197768068642,
+      "grad_norm": 1.945612907409668,
+      "learning_rate": 5.373448072580097e-06,
+      "loss": 0.1783,
+      "step": 37380
+    },
+    {
+      "epoch": 0.4921829762594094,
+      "grad_norm": 1.4023823738098145,
+      "learning_rate": 5.369197628555352e-06,
+      "loss": 0.195,
+      "step": 37400
+    },
+    {
+      "epoch": 0.49244617571195454,
+      "grad_norm": 2.0875327587127686,
+      "learning_rate": 5.36494691624995e-06,
+      "loss": 0.189,
+      "step": 37420
+    },
+    {
+      "epoch": 0.49270937516449964,
+      "grad_norm": 2.0927884578704834,
+      "learning_rate": 5.36069593875271e-06,
+      "loss": 0.1777,
+      "step": 37440
+    },
+    {
+      "epoch": 0.4929725746170448,
+      "grad_norm": 1.3830748796463013,
+      "learning_rate": 5.356444699152641e-06,
+      "loss": 0.2028,
+      "step": 37460
+    },
+    {
+      "epoch": 0.49323577406958996,
+      "grad_norm": 1.5777854919433594,
+      "learning_rate": 5.3521932005389434e-06,
+      "loss": 0.1874,
+      "step": 37480
+    },
+    {
+      "epoch": 0.49349897352213506,
+      "grad_norm": 2.379237174987793,
+      "learning_rate": 5.347941446001004e-06,
+      "loss": 0.1993,
+      "step": 37500
+    },
+    {
+      "epoch": 0.49349897352213506,
+      "eval_loss": 0.2187410295009613,
+      "eval_runtime": 3.9761,
+      "eval_samples_per_second": 4.024,
+      "eval_steps_per_second": 2.012,
+      "step": 37500
+    },
+    {
+      "epoch": 0.4937621729746802,
+      "grad_norm": 1.1107007265090942,
+      "learning_rate": 5.343689438628399e-06,
+      "loss": 0.183,
+      "step": 37520
+    },
+    {
+      "epoch": 0.49402537242722533,
+      "grad_norm": 1.877463698387146,
+      "learning_rate": 5.3394371815108855e-06,
+      "loss": 0.1884,
+      "step": 37540
+    },
+    {
+      "epoch": 0.4942885718797705,
+      "grad_norm": 1.6511403322219849,
+      "learning_rate": 5.335184677738403e-06,
+      "loss": 0.1674,
+      "step": 37560
+    },
+    {
+      "epoch": 0.49455177133231565,
+      "grad_norm": 1.3518986701965332,
+      "learning_rate": 5.330931930401071e-06,
+      "loss": 0.1864,
+      "step": 37580
+    },
+    {
+      "epoch": 0.49481497078486075,
+      "grad_norm": 1.5183402299880981,
+      "learning_rate": 5.326678942589185e-06,
+      "loss": 0.1776,
+      "step": 37600
+    },
+    {
+      "epoch": 0.4950781702374059,
+      "grad_norm": 1.6998218297958374,
+      "learning_rate": 5.322425717393215e-06,
+      "loss": 0.1865,
+      "step": 37620
+    },
+    {
+      "epoch": 0.49534136968995107,
+      "grad_norm": 1.7055150270462036,
+      "learning_rate": 5.318172257903802e-06,
+      "loss": 0.1923,
+      "step": 37640
+    },
+    {
+      "epoch": 0.4956045691424962,
+      "grad_norm": 1.4201116561889648,
+      "learning_rate": 5.3139185672117635e-06,
+      "loss": 0.1914,
+      "step": 37660
+    },
+    {
+      "epoch": 0.49586776859504134,
+      "grad_norm": 1.7442861795425415,
+      "learning_rate": 5.309664648408076e-06,
+      "loss": 0.1964,
+      "step": 37680
+    },
+    {
+      "epoch": 0.49613096804758644,
+      "grad_norm": 1.353569507598877,
+      "learning_rate": 5.305410504583891e-06,
+      "loss": 0.1903,
+      "step": 37700
+    },
+    {
+      "epoch": 0.4963941675001316,
+      "grad_norm": 1.7393383979797363,
+      "learning_rate": 5.301156138830516e-06,
+      "loss": 0.1838,
+      "step": 37720
+    },
+    {
+      "epoch": 0.49665736695267676,
+      "grad_norm": 1.2452112436294556,
+      "learning_rate": 5.296901554239423e-06,
+      "loss": 0.1957,
+      "step": 37740
+    },
+    {
+      "epoch": 0.49692056640522186,
+      "grad_norm": 1.1121643781661987,
+      "learning_rate": 5.292646753902243e-06,
+      "loss": 0.1852,
+      "step": 37760
+    },
+    {
+      "epoch": 0.497183765857767,
+      "grad_norm": 1.096800684928894,
+      "learning_rate": 5.2883917409107656e-06,
+      "loss": 0.1762,
+      "step": 37780
+    },
+    {
+      "epoch": 0.49744696531031213,
+      "grad_norm": 1.7458487749099731,
+      "learning_rate": 5.284136518356929e-06,
+      "loss": 0.1957,
+      "step": 37800
+    },
+    {
+      "epoch": 0.4977101647628573,
+      "grad_norm": 1.9648549556732178,
+      "learning_rate": 5.27988108933283e-06,
+      "loss": 0.1845,
+      "step": 37820
+    },
+    {
+      "epoch": 0.49797336421540245,
+      "grad_norm": 1.7593419551849365,
+      "learning_rate": 5.275625456930712e-06,
+      "loss": 0.1805,
+      "step": 37840
+    },
+    {
+      "epoch": 0.49823656366794755,
+      "grad_norm": 1.5812164545059204,
+      "learning_rate": 5.2713696242429705e-06,
+      "loss": 0.1877,
+      "step": 37860
+    },
+    {
+      "epoch": 0.4984997631204927,
+      "grad_norm": 1.706364631652832,
+      "learning_rate": 5.267113594362138e-06,
+      "loss": 0.1844,
+      "step": 37880
+    },
+    {
+      "epoch": 0.49876296257303787,
+      "grad_norm": 1.3167591094970703,
+      "learning_rate": 5.262857370380898e-06,
+      "loss": 0.1825,
+      "step": 37900
+    },
+    {
+      "epoch": 0.499026162025583,
+      "grad_norm": 1.5875823497772217,
+      "learning_rate": 5.258600955392071e-06,
+      "loss": 0.2041,
+      "step": 37920
+    },
+    {
+      "epoch": 0.49928936147812814,
+      "grad_norm": 1.6976323127746582,
+      "learning_rate": 5.254344352488622e-06,
+      "loss": 0.2026,
+      "step": 37940
+    },
+    {
+      "epoch": 0.49955256093067324,
+      "grad_norm": 1.5020610094070435,
+      "learning_rate": 5.250087564763645e-06,
+      "loss": 0.2003,
+      "step": 37960
+    },
+    {
+      "epoch": 0.4998157603832184,
+      "grad_norm": 1.3467730283737183,
+      "learning_rate": 5.2458305953103715e-06,
+      "loss": 0.1896,
+      "step": 37980
+    },
+    {
+      "epoch": 0.5000789598357636,
+      "grad_norm": 1.3582921028137207,
+      "learning_rate": 5.241573447222166e-06,
+      "loss": 0.1881,
+      "step": 38000
+    },
+    {
+      "epoch": 0.5000789598357636,
+      "eval_loss": 0.2194061130285263,
+      "eval_runtime": 4.4065,
+      "eval_samples_per_second": 3.631,
+      "eval_steps_per_second": 1.816,
+      "step": 38000
+    },
+    {
+      "epoch": 0.5003421592883087,
+      "grad_norm": 1.4528862237930298,
+      "learning_rate": 5.237316123592523e-06,
+      "loss": 0.1743,
+      "step": 38020
+    },
+    {
+      "epoch": 0.5006053587408538,
+      "grad_norm": 1.3000119924545288,
+      "learning_rate": 5.233058627515062e-06,
+      "loss": 0.1978,
+      "step": 38040
+    },
+    {
+      "epoch": 0.5008685581933989,
+      "grad_norm": 1.769835352897644,
+      "learning_rate": 5.228800962083529e-06,
+      "loss": 0.1836,
+      "step": 38060
+    },
+    {
+      "epoch": 0.5011317576459441,
+      "grad_norm": 1.3525722026824951,
+      "learning_rate": 5.224543130391796e-06,
+      "loss": 0.1893,
+      "step": 38080
+    },
+    {
+      "epoch": 0.5013949570984892,
+      "grad_norm": 1.3831286430358887,
+      "learning_rate": 5.2202851355338525e-06,
+      "loss": 0.1978,
+      "step": 38100
+    },
+    {
+      "epoch": 0.5016581565510344,
+      "grad_norm": 1.7284330129623413,
+      "learning_rate": 5.2160269806038055e-06,
+      "loss": 0.1811,
+      "step": 38120
+    },
+    {
+      "epoch": 0.5019213560035796,
+      "grad_norm": 1.9811046123504639,
+      "learning_rate": 5.211768668695882e-06,
+      "loss": 0.177,
+      "step": 38140
+    },
+    {
+      "epoch": 0.5021845554561246,
+      "grad_norm": 1.2399438619613647,
+      "learning_rate": 5.20751020290442e-06,
+      "loss": 0.1664,
+      "step": 38160
+    },
+    {
+      "epoch": 0.5024477549086698,
+      "grad_norm": 1.3734984397888184,
+      "learning_rate": 5.20325158632387e-06,
+      "loss": 0.1918,
+      "step": 38180
+    },
+    {
+      "epoch": 0.5027109543612149,
+      "grad_norm": 2.948352098464966,
+      "learning_rate": 5.198992822048796e-06,
+      "loss": 0.1941,
+      "step": 38200
+    },
+    {
+      "epoch": 0.5029741538137601,
+      "grad_norm": 1.7070094347000122,
+      "learning_rate": 5.194733913173863e-06,
+      "loss": 0.194,
+      "step": 38220
+    },
+    {
+      "epoch": 0.5032373532663053,
+      "grad_norm": 1.7377053499221802,
+      "learning_rate": 5.190474862793845e-06,
+      "loss": 0.1783,
+      "step": 38240
+    },
+    {
+      "epoch": 0.5035005527188503,
+      "grad_norm": 1.3507676124572754,
+      "learning_rate": 5.186215674003616e-06,
+      "loss": 0.1946,
+      "step": 38260
+    },
+    {
+      "epoch": 0.5037637521713955,
+      "grad_norm": 1.3682056665420532,
+      "learning_rate": 5.181956349898156e-06,
+      "loss": 0.1889,
+      "step": 38280
+    },
+    {
+      "epoch": 0.5040269516239406,
+      "grad_norm": 1.2678008079528809,
+      "learning_rate": 5.177696893572538e-06,
+      "loss": 0.1797,
+      "step": 38300
+    },
+    {
+      "epoch": 0.5042901510764858,
+      "grad_norm": 1.7389485836029053,
+      "learning_rate": 5.173437308121933e-06,
+      "loss": 0.1888,
+      "step": 38320
+    },
+    {
+      "epoch": 0.5045533505290309,
+      "grad_norm": 1.7758047580718994,
+      "learning_rate": 5.169177596641604e-06,
+      "loss": 0.1866,
+      "step": 38340
+    },
+    {
+      "epoch": 0.504816549981576,
+      "grad_norm": 1.3992506265640259,
+      "learning_rate": 5.164917762226908e-06,
+      "loss": 0.1698,
+      "step": 38360
+    },
+    {
+      "epoch": 0.5050797494341212,
+      "grad_norm": 2.007232189178467,
+      "learning_rate": 5.160657807973293e-06,
+      "loss": 0.1871,
+      "step": 38380
+    },
+    {
+      "epoch": 0.5053429488866663,
+      "grad_norm": 3.1283140182495117,
+      "learning_rate": 5.156397736976291e-06,
+      "loss": 0.1701,
+      "step": 38400
+    },
+    {
+      "epoch": 0.5056061483392115,
+      "grad_norm": 1.4590790271759033,
+      "learning_rate": 5.152137552331515e-06,
+      "loss": 0.1828,
+      "step": 38420
+    },
+    {
+      "epoch": 0.5058693477917566,
+      "grad_norm": 1.3053765296936035,
+      "learning_rate": 5.14787725713467e-06,
+      "loss": 0.1918,
+      "step": 38440
+    },
+    {
+      "epoch": 0.5061325472443017,
+      "grad_norm": 1.4517500400543213,
+      "learning_rate": 5.1436168544815345e-06,
+      "loss": 0.1941,
+      "step": 38460
+    },
+    {
+      "epoch": 0.5063957466968468,
+      "grad_norm": 1.6822638511657715,
+      "learning_rate": 5.13935634746797e-06,
+      "loss": 0.1754,
+      "step": 38480
+    },
+    {
+      "epoch": 0.506658946149392,
+      "grad_norm": 1.769856572151184,
+      "learning_rate": 5.135095739189905e-06,
+      "loss": 0.1894,
+      "step": 38500
+    },
+    {
+      "epoch": 0.506658946149392,
+      "eval_loss": 0.22107434272766113,
+      "eval_runtime": 3.5742,
+      "eval_samples_per_second": 4.477,
+      "eval_steps_per_second": 2.238,
+      "step": 38500
+    },
+    {
+      "epoch": 0.5069221456019372,
+      "grad_norm": 1.7754263877868652,
+      "learning_rate": 5.130835032743352e-06,
+      "loss": 0.2012,
+      "step": 38520
+    },
+    {
+      "epoch": 0.5071853450544823,
+      "grad_norm": 1.3758280277252197,
+      "learning_rate": 5.126574231224388e-06,
+      "loss": 0.1925,
+      "step": 38540
+    },
+    {
+      "epoch": 0.5074485445070275,
+      "grad_norm": 1.557976245880127,
+      "learning_rate": 5.122313337729164e-06,
+      "loss": 0.1829,
+      "step": 38560
+    },
+    {
+      "epoch": 0.5077117439595725,
+      "grad_norm": 1.4832298755645752,
+      "learning_rate": 5.118052355353892e-06,
+      "loss": 0.1776,
+      "step": 38580
+    },
+    {
+      "epoch": 0.5079749434121177,
+      "grad_norm": 1.385771632194519,
+      "learning_rate": 5.1137912871948535e-06,
+      "loss": 0.2011,
+      "step": 38600
+    },
+    {
+      "epoch": 0.5082381428646628,
+      "grad_norm": 1.8723502159118652,
+      "learning_rate": 5.10953013634839e-06,
+      "loss": 0.1964,
+      "step": 38620
+    },
+    {
+      "epoch": 0.508501342317208,
+      "grad_norm": 1.6367015838623047,
+      "learning_rate": 5.105268905910904e-06,
+      "loss": 0.1613,
+      "step": 38640
+    },
+    {
+      "epoch": 0.5087645417697532,
+      "grad_norm": 1.4688462018966675,
+      "learning_rate": 5.101007598978856e-06,
+      "loss": 0.1937,
+      "step": 38660
+    },
+    {
+      "epoch": 0.5090277412222982,
+      "grad_norm": 2.025383234024048,
+      "learning_rate": 5.096746218648761e-06,
+      "loss": 0.1972,
+      "step": 38680
+    },
+    {
+      "epoch": 0.5092909406748434,
+      "grad_norm": 1.4352595806121826,
+      "learning_rate": 5.092484768017188e-06,
+      "loss": 0.1979,
+      "step": 38700
+    },
+    {
+      "epoch": 0.5095541401273885,
+      "grad_norm": 1.2373008728027344,
+      "learning_rate": 5.088223250180755e-06,
+      "loss": 0.1871,
+      "step": 38720
+    },
+    {
+      "epoch": 0.5098173395799337,
+      "grad_norm": 1.8132870197296143,
+      "learning_rate": 5.083961668236136e-06,
+      "loss": 0.1637,
+      "step": 38740
+    },
+    {
+      "epoch": 0.5100805390324789,
+      "grad_norm": 1.3951473236083984,
+      "learning_rate": 5.079700025280041e-06,
+      "loss": 0.1873,
+      "step": 38760
+    },
+    {
+      "epoch": 0.5103437384850239,
+      "grad_norm": 1.4244937896728516,
+      "learning_rate": 5.075438324409233e-06,
+      "loss": 0.1756,
+      "step": 38780
+    },
+    {
+      "epoch": 0.5106069379375691,
+      "grad_norm": 1.521200180053711,
+      "learning_rate": 5.071176568720513e-06,
+      "loss": 0.184,
+      "step": 38800
+    },
+    {
+      "epoch": 0.5108701373901142,
+      "grad_norm": 1.1767172813415527,
+      "learning_rate": 5.066914761310724e-06,
+      "loss": 0.1696,
+      "step": 38820
+    },
+    {
+      "epoch": 0.5111333368426594,
+      "grad_norm": 1.3449183702468872,
+      "learning_rate": 5.062652905276741e-06,
+      "loss": 0.1815,
+      "step": 38840
+    },
+    {
+      "epoch": 0.5113965362952045,
+      "grad_norm": 1.4093093872070312,
+      "learning_rate": 5.058391003715484e-06,
+      "loss": 0.1823,
+      "step": 38860
+    },
+    {
+      "epoch": 0.5116597357477497,
+      "grad_norm": 1.8944716453552246,
+      "learning_rate": 5.0541290597238965e-06,
+      "loss": 0.1829,
+      "step": 38880
+    },
+    {
+      "epoch": 0.5119229352002947,
+      "grad_norm": 1.695489764213562,
+      "learning_rate": 5.04986707639896e-06,
+      "loss": 0.1969,
+      "step": 38900
+    },
+    {
+      "epoch": 0.5121861346528399,
+      "grad_norm": 2.140026092529297,
+      "learning_rate": 5.04560505683768e-06,
+      "loss": 0.1984,
+      "step": 38920
+    },
+    {
+      "epoch": 0.5124493341053851,
+      "grad_norm": 1.9395867586135864,
+      "learning_rate": 5.0413430041370915e-06,
+      "loss": 0.1702,
+      "step": 38940
+    },
+    {
+      "epoch": 0.5127125335579302,
+      "grad_norm": 1.3901492357254028,
+      "learning_rate": 5.037080921394248e-06,
+      "loss": 0.1669,
+      "step": 38960
+    },
+    {
+      "epoch": 0.5129757330104754,
+      "grad_norm": 1.8665878772735596,
+      "learning_rate": 5.032818811706235e-06,
+      "loss": 0.182,
+      "step": 38980
+    },
+    {
+      "epoch": 0.5132389324630204,
+      "grad_norm": 1.4569926261901855,
+      "learning_rate": 5.028556678170147e-06,
+      "loss": 0.1763,
+      "step": 39000
+    },
+    {
+      "epoch": 0.5132389324630204,
+      "eval_loss": 0.21912354230880737,
+      "eval_runtime": 4.064,
+      "eval_samples_per_second": 3.937,
+      "eval_steps_per_second": 1.968,
+      "step": 39000
+    },
+    {
+      "epoch": 0.5135021319155656,
+      "grad_norm": 1.8636062145233154,
+      "learning_rate": 5.0242945238831005e-06,
+      "loss": 0.1875,
+      "step": 39020
+    },
+    {
+      "epoch": 0.5137653313681108,
+      "grad_norm": 1.5546830892562866,
+      "learning_rate": 5.020032351942228e-06,
+      "loss": 0.1851,
+      "step": 39040
+    },
+    {
+      "epoch": 0.5140285308206559,
+      "grad_norm": 1.5957520008087158,
+      "learning_rate": 5.015770165444673e-06,
+      "loss": 0.1976,
+      "step": 39060
+    },
+    {
+      "epoch": 0.5142917302732011,
+      "grad_norm": 1.2627102136611938,
+      "learning_rate": 5.011507967487591e-06,
+      "loss": 0.1814,
+      "step": 39080
+    },
+    {
+      "epoch": 0.5145549297257461,
+      "grad_norm": 1.8712400197982788,
+      "learning_rate": 5.007245761168147e-06,
+      "loss": 0.1825,
+      "step": 39100
+    },
+    {
+      "epoch": 0.5148181291782913,
+      "grad_norm": 1.4492478370666504,
+      "learning_rate": 5.002983549583505e-06,
+      "loss": 0.1977,
+      "step": 39120
+    },
+    {
+      "epoch": 0.5150813286308364,
+      "grad_norm": 2.0389404296875,
+      "learning_rate": 4.998721335830842e-06,
+      "loss": 0.1634,
+      "step": 39140
+    },
+    {
+      "epoch": 0.5153445280833816,
+      "grad_norm": 1.4813175201416016,
+      "learning_rate": 4.9944591230073305e-06,
+      "loss": 0.1919,
+      "step": 39160
+    },
+    {
+      "epoch": 0.5156077275359268,
+      "grad_norm": 1.3129205703735352,
+      "learning_rate": 4.990196914210145e-06,
+      "loss": 0.2139,
+      "step": 39180
+    },
+    {
+      "epoch": 0.5158709269884718,
+      "grad_norm": 1.4541630744934082,
+      "learning_rate": 4.985934712536458e-06,
+      "loss": 0.1746,
+      "step": 39200
+    },
+    {
+      "epoch": 0.516134126441017,
+      "grad_norm": 1.4930351972579956,
+      "learning_rate": 4.981672521083434e-06,
+      "loss": 0.1753,
+      "step": 39220
+    },
+    {
+      "epoch": 0.5163973258935621,
+      "grad_norm": 1.3711662292480469,
+      "learning_rate": 4.977410342948231e-06,
+      "loss": 0.1723,
+      "step": 39240
+    },
+    {
+      "epoch": 0.5166605253461073,
+      "grad_norm": 1.5443474054336548,
+      "learning_rate": 4.973148181227998e-06,
+      "loss": 0.1717,
+      "step": 39260
+    },
+    {
+      "epoch": 0.5169237247986525,
+      "grad_norm": 1.4532698392868042,
+      "learning_rate": 4.96888603901987e-06,
+      "loss": 0.1999,
+      "step": 39280
+    },
+    {
+      "epoch": 0.5171869242511976,
+      "grad_norm": 1.4252219200134277,
+      "learning_rate": 4.964623919420974e-06,
+      "loss": 0.1733,
+      "step": 39300
+    },
+    {
+      "epoch": 0.5174501237037427,
+      "grad_norm": 1.6706379652023315,
+      "learning_rate": 4.960361825528414e-06,
+      "loss": 0.1788,
+      "step": 39320
+    },
+    {
+      "epoch": 0.5177133231562878,
+      "grad_norm": 1.7905189990997314,
+      "learning_rate": 4.956099760439274e-06,
+      "loss": 0.1856,
+      "step": 39340
+    },
+    {
+      "epoch": 0.517976522608833,
+      "grad_norm": 1.1484737396240234,
+      "learning_rate": 4.951837727250626e-06,
+      "loss": 0.1838,
+      "step": 39360
+    },
+    {
+      "epoch": 0.5182397220613781,
+      "grad_norm": 1.407689094543457,
+      "learning_rate": 4.9475757290595085e-06,
+      "loss": 0.1741,
+      "step": 39380
+    },
+    {
+      "epoch": 0.5185029215139233,
+      "grad_norm": 1.3493037223815918,
+      "learning_rate": 4.943313768962943e-06,
+      "loss": 0.2054,
+      "step": 39400
+    },
+    {
+      "epoch": 0.5187661209664683,
+      "grad_norm": 1.2976672649383545,
+      "learning_rate": 4.939051850057918e-06,
+      "loss": 0.1628,
+      "step": 39420
+    },
+    {
+      "epoch": 0.5190293204190135,
+      "grad_norm": 1.8429856300354004,
+      "learning_rate": 4.934789975441394e-06,
+      "loss": 0.1678,
+      "step": 39440
+    },
+    {
+      "epoch": 0.5192925198715587,
+      "grad_norm": 1.4335477352142334,
+      "learning_rate": 4.9305281482103e-06,
+      "loss": 0.1905,
+      "step": 39460
+    },
+    {
+      "epoch": 0.5195557193241038,
+      "grad_norm": 1.47794771194458,
+      "learning_rate": 4.92626637146153e-06,
+      "loss": 0.199,
+      "step": 39480
+    },
+    {
+      "epoch": 0.519818918776649,
+      "grad_norm": 1.4035592079162598,
+      "learning_rate": 4.922004648291937e-06,
+      "loss": 0.1749,
+      "step": 39500
+    },
+    {
+      "epoch": 0.519818918776649,
+      "eval_loss": 0.21317389607429504,
+      "eval_runtime": 4.1385,
+      "eval_samples_per_second": 3.866,
+      "eval_steps_per_second": 1.933,
+      "step": 39500
+    },
+    {
+      "epoch": 0.520082118229194,
+      "grad_norm": 2.4462828636169434,
+      "learning_rate": 4.917742981798345e-06,
+      "loss": 0.173,
+      "step": 39520
+    },
+    {
+      "epoch": 0.5203453176817392,
+      "grad_norm": 1.8106077909469604,
+      "learning_rate": 4.913481375077525e-06,
+      "loss": 0.2031,
+      "step": 39540
+    },
+    {
+      "epoch": 0.5206085171342844,
+      "grad_norm": 1.1889911890029907,
+      "learning_rate": 4.909219831226217e-06,
+      "loss": 0.186,
+      "step": 39560
+    },
+    {
+      "epoch": 0.5208717165868295,
+      "grad_norm": 1.3789622783660889,
+      "learning_rate": 4.904958353341103e-06,
+      "loss": 0.1999,
+      "step": 39580
+    },
+    {
+      "epoch": 0.5211349160393747,
+      "grad_norm": 1.479336142539978,
+      "learning_rate": 4.900696944518825e-06,
+      "loss": 0.216,
+      "step": 39600
+    },
+    {
+      "epoch": 0.5213981154919197,
+      "grad_norm": 1.1836003065109253,
+      "learning_rate": 4.896435607855973e-06,
+      "loss": 0.196,
+      "step": 39620
+    },
+    {
+      "epoch": 0.5216613149444649,
+      "grad_norm": 1.3221081495285034,
+      "learning_rate": 4.8921743464490836e-06,
+      "loss": 0.1906,
+      "step": 39640
+    },
+    {
+      "epoch": 0.52192451439701,
+      "grad_norm": 1.8187506198883057,
+      "learning_rate": 4.88791316339464e-06,
+      "loss": 0.1819,
+      "step": 39660
+    },
+    {
+      "epoch": 0.5221877138495552,
+      "grad_norm": 1.5414676666259766,
+      "learning_rate": 4.883652061789067e-06,
+      "loss": 0.1832,
+      "step": 39680
+    },
+    {
+      "epoch": 0.5224509133021004,
+      "grad_norm": 1.095396876335144,
+      "learning_rate": 4.879391044728731e-06,
+      "loss": 0.182,
+      "step": 39700
+    },
+    {
+      "epoch": 0.5227141127546455,
+      "grad_norm": 1.5563472509384155,
+      "learning_rate": 4.875130115309938e-06,
+      "loss": 0.184,
+      "step": 39720
+    },
+    {
+      "epoch": 0.5229773122071906,
+      "grad_norm": 1.7773133516311646,
+      "learning_rate": 4.87086927662893e-06,
+      "loss": 0.2063,
+      "step": 39740
+    },
+    {
+      "epoch": 0.5232405116597357,
+      "grad_norm": 2.158665180206299,
+      "learning_rate": 4.866608531781878e-06,
+      "loss": 0.2049,
+      "step": 39760
+    },
+    {
+      "epoch": 0.5235037111122809,
+      "grad_norm": 1.7402114868164062,
+      "learning_rate": 4.8623478838648945e-06,
+      "loss": 0.1892,
+      "step": 39780
+    },
+    {
+      "epoch": 0.523766910564826,
+      "grad_norm": 1.5573785305023193,
+      "learning_rate": 4.858087335974011e-06,
+      "loss": 0.208,
+      "step": 39800
+    },
+    {
+      "epoch": 0.5240301100173712,
+      "grad_norm": 1.1120315790176392,
+      "learning_rate": 4.8538268912051965e-06,
+      "loss": 0.1974,
+      "step": 39820
+    },
+    {
+      "epoch": 0.5242933094699163,
+      "grad_norm": 1.9333202838897705,
+      "learning_rate": 4.849566552654338e-06,
+      "loss": 0.2018,
+      "step": 39840
+    },
+    {
+      "epoch": 0.5245565089224614,
+      "grad_norm": 1.6155110597610474,
+      "learning_rate": 4.845306323417246e-06,
+      "loss": 0.1648,
+      "step": 39860
+    },
+    {
+      "epoch": 0.5248197083750066,
+      "grad_norm": 1.3805612325668335,
+      "learning_rate": 4.841046206589656e-06,
+      "loss": 0.1696,
+      "step": 39880
+    },
+    {
+      "epoch": 0.5250829078275517,
+      "grad_norm": 1.6295537948608398,
+      "learning_rate": 4.836786205267215e-06,
+      "loss": 0.2098,
+      "step": 39900
+    },
+    {
+      "epoch": 0.5253461072800969,
+      "grad_norm": 1.6805764436721802,
+      "learning_rate": 4.832526322545494e-06,
+      "loss": 0.2022,
+      "step": 39920
+    },
+    {
+      "epoch": 0.525609306732642,
+      "grad_norm": 1.6541625261306763,
+      "learning_rate": 4.8282665615199725e-06,
+      "loss": 0.1835,
+      "step": 39940
+    },
+    {
+      "epoch": 0.5258725061851871,
+      "grad_norm": 1.5509734153747559,
+      "learning_rate": 4.824006925286039e-06,
+      "loss": 0.1808,
+      "step": 39960
+    },
+    {
+      "epoch": 0.5261357056377323,
+      "grad_norm": 1.4882410764694214,
+      "learning_rate": 4.819747416939001e-06,
+      "loss": 0.1627,
+      "step": 39980
+    },
+    {
+      "epoch": 0.5263989050902774,
+      "grad_norm": 1.2531769275665283,
+      "learning_rate": 4.815488039574061e-06,
+      "loss": 0.184,
+      "step": 40000
+    },
+    {
+      "epoch": 0.5263989050902774,
+      "eval_loss": 0.20719100534915924,
+      "eval_runtime": 4.1022,
+      "eval_samples_per_second": 3.9,
+      "eval_steps_per_second": 1.95,
+      "step": 40000
+    },
+    {
+      "epoch": 0.5266621045428226,
+      "grad_norm": 1.5981402397155762,
+      "learning_rate": 4.811228796286339e-06,
+      "loss": 0.1827,
+      "step": 40020
+    },
+    {
+      "epoch": 0.5269253039953677,
+      "grad_norm": 1.3127453327178955,
+      "learning_rate": 4.806969690170846e-06,
+      "loss": 0.1821,
+      "step": 40040
+    },
+    {
+      "epoch": 0.5271885034479128,
+      "grad_norm": 1.1714023351669312,
+      "learning_rate": 4.8027107243225e-06,
+      "loss": 0.1945,
+      "step": 40060
+    },
+    {
+      "epoch": 0.527451702900458,
+      "grad_norm": 1.3002939224243164,
+      "learning_rate": 4.798451901836116e-06,
+      "loss": 0.1902,
+      "step": 40080
+    },
+    {
+      "epoch": 0.5277149023530031,
+      "grad_norm": 1.4868640899658203,
+      "learning_rate": 4.794193225806405e-06,
+      "loss": 0.1846,
+      "step": 40100
+    },
+    {
+      "epoch": 0.5279781018055483,
+      "grad_norm": 1.2027915716171265,
+      "learning_rate": 4.789934699327966e-06,
+      "loss": 0.1766,
+      "step": 40120
+    },
+    {
+      "epoch": 0.5282413012580934,
+      "grad_norm": 1.6551827192306519,
+      "learning_rate": 4.785676325495302e-06,
+      "loss": 0.1862,
+      "step": 40140
+    },
+    {
+      "epoch": 0.5285045007106385,
+      "grad_norm": 1.4043769836425781,
+      "learning_rate": 4.781418107402789e-06,
+      "loss": 0.1963,
+      "step": 40160
+    },
+    {
+      "epoch": 0.5287677001631836,
+      "grad_norm": 1.394041657447815,
+      "learning_rate": 4.777160048144706e-06,
+      "loss": 0.1878,
+      "step": 40180
+    },
+    {
+      "epoch": 0.5290308996157288,
+      "grad_norm": 1.530765175819397,
+      "learning_rate": 4.772902150815202e-06,
+      "loss": 0.1666,
+      "step": 40200
+    },
+    {
+      "epoch": 0.529294099068274,
+      "grad_norm": 1.932599425315857,
+      "learning_rate": 4.768644418508318e-06,
+      "loss": 0.1828,
+      "step": 40220
+    },
+    {
+      "epoch": 0.5295572985208191,
+      "grad_norm": 1.5278692245483398,
+      "learning_rate": 4.764386854317972e-06,
+      "loss": 0.1811,
+      "step": 40240
+    },
+    {
+      "epoch": 0.5298204979733642,
+      "grad_norm": 1.5748794078826904,
+      "learning_rate": 4.76012946133796e-06,
+      "loss": 0.1829,
+      "step": 40260
+    },
+    {
+      "epoch": 0.5300836974259093,
+      "grad_norm": 1.2589044570922852,
+      "learning_rate": 4.755872242661953e-06,
+      "loss": 0.177,
+      "step": 40280
+    },
+    {
+      "epoch": 0.5303468968784545,
+      "grad_norm": 1.7446081638336182,
+      "learning_rate": 4.751615201383496e-06,
+      "loss": 0.18,
+      "step": 40300
+    },
+    {
+      "epoch": 0.5306100963309996,
+      "grad_norm": 1.3446251153945923,
+      "learning_rate": 4.747358340596004e-06,
+      "loss": 0.1871,
+      "step": 40320
+    },
+    {
+      "epoch": 0.5308732957835448,
+      "grad_norm": 2.092999219894409,
+      "learning_rate": 4.743101663392764e-06,
+      "loss": 0.1884,
+      "step": 40340
+    },
+    {
+      "epoch": 0.5311364952360899,
+      "grad_norm": 1.6954139471054077,
+      "learning_rate": 4.7388451728669274e-06,
+      "loss": 0.1857,
+      "step": 40360
+    },
+    {
+      "epoch": 0.531399694688635,
+      "grad_norm": 1.6433746814727783,
+      "learning_rate": 4.734588872111507e-06,
+      "loss": 0.217,
+      "step": 40380
+    },
+    {
+      "epoch": 0.5316628941411802,
+      "grad_norm": 1.7384026050567627,
+      "learning_rate": 4.730332764219382e-06,
+      "loss": 0.182,
+      "step": 40400
+    },
+    {
+      "epoch": 0.5319260935937253,
+      "grad_norm": 1.335483193397522,
+      "learning_rate": 4.726076852283289e-06,
+      "loss": 0.1875,
+      "step": 40420
+    },
+    {
+      "epoch": 0.5321892930462705,
+      "grad_norm": 1.6717461347579956,
+      "learning_rate": 4.721821139395827e-06,
+      "loss": 0.1701,
+      "step": 40440
+    },
+    {
+      "epoch": 0.5324524924988157,
+      "grad_norm": 1.766654133796692,
+      "learning_rate": 4.717565628649443e-06,
+      "loss": 0.1899,
+      "step": 40460
+    },
+    {
+      "epoch": 0.5327156919513607,
+      "grad_norm": 1.5605034828186035,
+      "learning_rate": 4.713310323136442e-06,
+      "loss": 0.1813,
+      "step": 40480
+    },
+    {
+      "epoch": 0.5329788914039059,
+      "grad_norm": 1.762020230293274,
+      "learning_rate": 4.7090552259489785e-06,
+      "loss": 0.1864,
+      "step": 40500
+    },
+    {
+      "epoch": 0.5329788914039059,
+      "eval_loss": 0.2077658623456955,
+      "eval_runtime": 3.8571,
+      "eval_samples_per_second": 4.148,
+      "eval_steps_per_second": 2.074,
+      "step": 40500
+    },
+    {
+      "epoch": 0.533242090856451,
+      "grad_norm": 1.4025185108184814,
+      "learning_rate": 4.704800340179055e-06,
+      "loss": 0.1754,
+      "step": 40520
+    },
+    {
+      "epoch": 0.5335052903089962,
+      "grad_norm": 1.2111016511917114,
+      "learning_rate": 4.7005456689185244e-06,
+      "loss": 0.163,
+      "step": 40540
+    },
+    {
+      "epoch": 0.5337684897615413,
+      "grad_norm": 1.7409456968307495,
+      "learning_rate": 4.696291215259078e-06,
+      "loss": 0.1755,
+      "step": 40560
+    },
+    {
+      "epoch": 0.5340316892140864,
+      "grad_norm": 2.062086582183838,
+      "learning_rate": 4.692036982292248e-06,
+      "loss": 0.2146,
+      "step": 40580
+    },
+    {
+      "epoch": 0.5342948886666316,
+      "grad_norm": 1.4869282245635986,
+      "learning_rate": 4.687782973109416e-06,
+      "loss": 0.1707,
+      "step": 40600
+    },
+    {
+      "epoch": 0.5345580881191767,
+      "grad_norm": 1.3704824447631836,
+      "learning_rate": 4.683529190801791e-06,
+      "loss": 0.1723,
+      "step": 40620
+    },
+    {
+      "epoch": 0.5348212875717219,
+      "grad_norm": 1.4955403804779053,
+      "learning_rate": 4.67927563846042e-06,
+      "loss": 0.1761,
+      "step": 40640
+    },
+    {
+      "epoch": 0.535084487024267,
+      "grad_norm": 1.2220302820205688,
+      "learning_rate": 4.6750223191761865e-06,
+      "loss": 0.1916,
+      "step": 40660
+    },
+    {
+      "epoch": 0.5353476864768121,
+      "grad_norm": 1.873802661895752,
+      "learning_rate": 4.6707692360397996e-06,
+      "loss": 0.1631,
+      "step": 40680
+    },
+    {
+      "epoch": 0.5356108859293572,
+      "grad_norm": 1.1960867643356323,
+      "learning_rate": 4.666516392141801e-06,
+      "loss": 0.1804,
+      "step": 40700
+    },
+    {
+      "epoch": 0.5358740853819024,
+      "grad_norm": 2.2512047290802,
+      "learning_rate": 4.662263790572555e-06,
+      "loss": 0.1885,
+      "step": 40720
+    },
+    {
+      "epoch": 0.5361372848344476,
+      "grad_norm": 1.250015377998352,
+      "learning_rate": 4.65801143442225e-06,
+      "loss": 0.1794,
+      "step": 40740
+    },
+    {
+      "epoch": 0.5364004842869927,
+      "grad_norm": 1.5824809074401855,
+      "learning_rate": 4.653759326780902e-06,
+      "loss": 0.1878,
+      "step": 40760
+    },
+    {
+      "epoch": 0.5366636837395378,
+      "grad_norm": 1.4592108726501465,
+      "learning_rate": 4.649507470738336e-06,
+      "loss": 0.1836,
+      "step": 40780
+    },
+    {
+      "epoch": 0.5369268831920829,
+      "grad_norm": 1.4382795095443726,
+      "learning_rate": 4.645255869384207e-06,
+      "loss": 0.1941,
+      "step": 40800
+    },
+    {
+      "epoch": 0.5371900826446281,
+      "grad_norm": 2.147789478302002,
+      "learning_rate": 4.6410045258079715e-06,
+      "loss": 0.1773,
+      "step": 40820
+    },
+    {
+      "epoch": 0.5374532820971732,
+      "grad_norm": 1.294313669204712,
+      "learning_rate": 4.636753443098907e-06,
+      "loss": 0.1868,
+      "step": 40840
+    },
+    {
+      "epoch": 0.5377164815497184,
+      "grad_norm": 1.2370857000350952,
+      "learning_rate": 4.632502624346102e-06,
+      "loss": 0.1828,
+      "step": 40860
+    },
+    {
+      "epoch": 0.5379796810022636,
+      "grad_norm": 1.7648144960403442,
+      "learning_rate": 4.6282520726384486e-06,
+      "loss": 0.1919,
+      "step": 40880
+    },
+    {
+      "epoch": 0.5382428804548086,
+      "grad_norm": 1.6009800434112549,
+      "learning_rate": 4.624001791064646e-06,
+      "loss": 0.1793,
+      "step": 40900
+    },
+    {
+      "epoch": 0.5385060799073538,
+      "grad_norm": 1.4022279977798462,
+      "learning_rate": 4.619751782713201e-06,
+      "loss": 0.1768,
+      "step": 40920
+    },
+    {
+      "epoch": 0.5387692793598989,
+      "grad_norm": 1.3161512613296509,
+      "learning_rate": 4.615502050672416e-06,
+      "loss": 0.2053,
+      "step": 40940
+    },
+    {
+      "epoch": 0.5390324788124441,
+      "grad_norm": 1.264475703239441,
+      "learning_rate": 4.611252598030399e-06,
+      "loss": 0.1826,
+      "step": 40960
+    },
+    {
+      "epoch": 0.5392956782649893,
+      "grad_norm": 1.3833750486373901,
+      "learning_rate": 4.607003427875051e-06,
+      "loss": 0.1852,
+      "step": 40980
+    },
+    {
+      "epoch": 0.5395588777175343,
+      "grad_norm": 1.686737298965454,
+      "learning_rate": 4.602754543294064e-06,
+      "loss": 0.1801,
+      "step": 41000
+    },
+    {
+      "epoch": 0.5395588777175343,
+      "eval_loss": 0.22013933956623077,
+      "eval_runtime": 3.9018,
+      "eval_samples_per_second": 4.101,
+      "eval_steps_per_second": 2.05,
+      "step": 41000
+    },
+    {
+      "epoch": 0.5398220771700795,
+      "grad_norm": 1.8988898992538452,
+      "learning_rate": 4.598505947374931e-06,
+      "loss": 0.1909,
+      "step": 41020
+    },
+    {
+      "epoch": 0.5400852766226246,
+      "grad_norm": 1.6533387899398804,
+      "learning_rate": 4.59425764320493e-06,
+      "loss": 0.1706,
+      "step": 41040
+    },
+    {
+      "epoch": 0.5403484760751698,
+      "grad_norm": 1.4029088020324707,
+      "learning_rate": 4.590009633871128e-06,
+      "loss": 0.1923,
+      "step": 41060
+    },
+    {
+      "epoch": 0.5406116755277149,
+      "grad_norm": 1.5201548337936401,
+      "learning_rate": 4.585761922460379e-06,
+      "loss": 0.1812,
+      "step": 41080
+    },
+    {
+      "epoch": 0.54087487498026,
+      "grad_norm": 1.6940970420837402,
+      "learning_rate": 4.581514512059315e-06,
+      "loss": 0.1741,
+      "step": 41100
+    },
+    {
+      "epoch": 0.5411380744328051,
+      "grad_norm": 1.4267889261245728,
+      "learning_rate": 4.5772674057543595e-06,
+      "loss": 0.1896,
+      "step": 41120
+    },
+    {
+      "epoch": 0.5414012738853503,
+      "grad_norm": 1.1731345653533936,
+      "learning_rate": 4.573020606631706e-06,
+      "loss": 0.193,
+      "step": 41140
+    },
+    {
+      "epoch": 0.5416644733378955,
+      "grad_norm": 1.2213419675827026,
+      "learning_rate": 4.568774117777326e-06,
+      "loss": 0.1759,
+      "step": 41160
+    },
+    {
+      "epoch": 0.5419276727904406,
+      "grad_norm": 1.6738967895507812,
+      "learning_rate": 4.564527942276971e-06,
+      "loss": 0.1822,
+      "step": 41180
+    },
+    {
+      "epoch": 0.5421908722429858,
+      "grad_norm": 1.1962687969207764,
+      "learning_rate": 4.560282083216158e-06,
+      "loss": 0.1935,
+      "step": 41200
+    },
+    {
+      "epoch": 0.5424540716955308,
+      "grad_norm": 1.1292619705200195,
+      "learning_rate": 4.556036543680182e-06,
+      "loss": 0.1911,
+      "step": 41220
+    },
+    {
+      "epoch": 0.542717271148076,
+      "grad_norm": 1.4278331995010376,
+      "learning_rate": 4.551791326754097e-06,
+      "loss": 0.1753,
+      "step": 41240
+    },
+    {
+      "epoch": 0.5429804706006212,
+      "grad_norm": 1.2059117555618286,
+      "learning_rate": 4.547546435522725e-06,
+      "loss": 0.193,
+      "step": 41260
+    },
+    {
+      "epoch": 0.5432436700531663,
+      "grad_norm": 1.487465500831604,
+      "learning_rate": 4.543301873070659e-06,
+      "loss": 0.1843,
+      "step": 41280
+    },
+    {
+      "epoch": 0.5435068695057115,
+      "grad_norm": 1.076802134513855,
+      "learning_rate": 4.539057642482242e-06,
+      "loss": 0.1885,
+      "step": 41300
+    },
+    {
+      "epoch": 0.5437700689582565,
+      "grad_norm": 1.801487922668457,
+      "learning_rate": 4.534813746841583e-06,
+      "loss": 0.1722,
+      "step": 41320
+    },
+    {
+      "epoch": 0.5440332684108017,
+      "grad_norm": 1.940256953239441,
+      "learning_rate": 4.530570189232546e-06,
+      "loss": 0.2002,
+      "step": 41340
+    },
+    {
+      "epoch": 0.5442964678633468,
+      "grad_norm": 1.7318363189697266,
+      "learning_rate": 4.526326972738747e-06,
+      "loss": 0.1875,
+      "step": 41360
+    },
+    {
+      "epoch": 0.544559667315892,
+      "grad_norm": 1.8844788074493408,
+      "learning_rate": 4.522084100443559e-06,
+      "loss": 0.1762,
+      "step": 41380
+    },
+    {
+      "epoch": 0.5448228667684372,
+      "grad_norm": 1.380464792251587,
+      "learning_rate": 4.517841575430098e-06,
+      "loss": 0.1723,
+      "step": 41400
+    },
+    {
+      "epoch": 0.5450860662209822,
+      "grad_norm": 1.5735033750534058,
+      "learning_rate": 4.513599400781235e-06,
+      "loss": 0.1936,
+      "step": 41420
+    },
+    {
+      "epoch": 0.5453492656735274,
+      "grad_norm": 1.3236199617385864,
+      "learning_rate": 4.509357579579582e-06,
+      "loss": 0.1721,
+      "step": 41440
+    },
+    {
+      "epoch": 0.5456124651260725,
+      "grad_norm": 1.6912643909454346,
+      "learning_rate": 4.505116114907493e-06,
+      "loss": 0.1851,
+      "step": 41460
+    },
+    {
+      "epoch": 0.5458756645786177,
+      "grad_norm": 1.8328466415405273,
+      "learning_rate": 4.500875009847069e-06,
+      "loss": 0.2003,
+      "step": 41480
+    },
+    {
+      "epoch": 0.5461388640311629,
+      "grad_norm": 1.964607834815979,
+      "learning_rate": 4.4966342674801425e-06,
+      "loss": 0.1784,
+      "step": 41500
+    },
+    {
+      "epoch": 0.5461388640311629,
+      "eval_loss": 0.22275885939598083,
+      "eval_runtime": 4.1004,
+      "eval_samples_per_second": 3.902,
+      "eval_steps_per_second": 1.951,
+      "step": 41500
+    },
+    {
+      "epoch": 0.5464020634837079,
+      "grad_norm": 1.83335280418396,
+      "learning_rate": 4.492393890888286e-06,
+      "loss": 0.182,
+      "step": 41520
+    },
+    {
+      "epoch": 0.5466652629362531,
+      "grad_norm": 1.3095813989639282,
+      "learning_rate": 4.4881538831528075e-06,
+      "loss": 0.1876,
+      "step": 41540
+    },
+    {
+      "epoch": 0.5469284623887982,
+      "grad_norm": 1.541033148765564,
+      "learning_rate": 4.483914247354743e-06,
+      "loss": 0.1789,
+      "step": 41560
+    },
+    {
+      "epoch": 0.5471916618413434,
+      "grad_norm": 1.7996163368225098,
+      "learning_rate": 4.479674986574864e-06,
+      "loss": 0.1826,
+      "step": 41580
+    },
+    {
+      "epoch": 0.5474548612938885,
+      "grad_norm": 1.6412596702575684,
+      "learning_rate": 4.475436103893664e-06,
+      "loss": 0.1878,
+      "step": 41600
+    },
+    {
+      "epoch": 0.5477180607464337,
+      "grad_norm": 1.4943265914916992,
+      "learning_rate": 4.4711976023913615e-06,
+      "loss": 0.1779,
+      "step": 41620
+    },
+    {
+      "epoch": 0.5479812601989787,
+      "grad_norm": 1.509543538093567,
+      "learning_rate": 4.466959485147906e-06,
+      "loss": 0.1644,
+      "step": 41640
+    },
+    {
+      "epoch": 0.5482444596515239,
+      "grad_norm": 1.5921076536178589,
+      "learning_rate": 4.462721755242956e-06,
+      "loss": 0.1902,
+      "step": 41660
+    },
+    {
+      "epoch": 0.5485076591040691,
+      "grad_norm": 1.3468427658081055,
+      "learning_rate": 4.4584844157559e-06,
+      "loss": 0.1797,
+      "step": 41680
+    },
+    {
+      "epoch": 0.5487708585566142,
+      "grad_norm": 1.3191331624984741,
+      "learning_rate": 4.454247469765835e-06,
+      "loss": 0.1838,
+      "step": 41700
+    },
+    {
+      "epoch": 0.5490340580091594,
+      "grad_norm": 2.3382019996643066,
+      "learning_rate": 4.450010920351573e-06,
+      "loss": 0.1648,
+      "step": 41720
+    },
+    {
+      "epoch": 0.5492972574617044,
+      "grad_norm": 1.641099452972412,
+      "learning_rate": 4.445774770591642e-06,
+      "loss": 0.1777,
+      "step": 41740
+    },
+    {
+      "epoch": 0.5495604569142496,
+      "grad_norm": 1.3900223970413208,
+      "learning_rate": 4.441539023564276e-06,
+      "loss": 0.1927,
+      "step": 41760
+    },
+    {
+      "epoch": 0.5498236563667948,
+      "grad_norm": 1.5319808721542358,
+      "learning_rate": 4.437303682347415e-06,
+      "loss": 0.1895,
+      "step": 41780
+    },
+    {
+      "epoch": 0.5500868558193399,
+      "grad_norm": 1.1724438667297363,
+      "learning_rate": 4.433068750018711e-06,
+      "loss": 0.1767,
+      "step": 41800
+    },
+    {
+      "epoch": 0.5503500552718851,
+      "grad_norm": 1.2725435495376587,
+      "learning_rate": 4.428834229655509e-06,
+      "loss": 0.1632,
+      "step": 41820
+    },
+    {
+      "epoch": 0.5506132547244301,
+      "grad_norm": 1.4931374788284302,
+      "learning_rate": 4.424600124334866e-06,
+      "loss": 0.185,
+      "step": 41840
+    },
+    {
+      "epoch": 0.5508764541769753,
+      "grad_norm": 2.1885902881622314,
+      "learning_rate": 4.420366437133526e-06,
+      "loss": 0.1897,
+      "step": 41860
+    },
+    {
+      "epoch": 0.5511396536295204,
+      "grad_norm": 1.9603490829467773,
+      "learning_rate": 4.416133171127936e-06,
+      "loss": 0.1949,
+      "step": 41880
+    },
+    {
+      "epoch": 0.5514028530820656,
+      "grad_norm": 1.73301362991333,
+      "learning_rate": 4.411900329394236e-06,
+      "loss": 0.2018,
+      "step": 41900
+    },
+    {
+      "epoch": 0.5516660525346108,
+      "grad_norm": 1.3896695375442505,
+      "learning_rate": 4.407667915008257e-06,
+      "loss": 0.1949,
+      "step": 41920
+    },
+    {
+      "epoch": 0.5519292519871559,
+      "grad_norm": 1.5166871547698975,
+      "learning_rate": 4.4034359310455205e-06,
+      "loss": 0.187,
+      "step": 41940
+    },
+    {
+      "epoch": 0.552192451439701,
+      "grad_norm": 1.2612345218658447,
+      "learning_rate": 4.399204380581233e-06,
+      "loss": 0.1845,
+      "step": 41960
+    },
+    {
+      "epoch": 0.5524556508922461,
+      "grad_norm": 1.9830892086029053,
+      "learning_rate": 4.394973266690287e-06,
+      "loss": 0.1648,
+      "step": 41980
+    },
+    {
+      "epoch": 0.5527188503447913,
+      "grad_norm": 1.5724226236343384,
+      "learning_rate": 4.390742592447259e-06,
+      "loss": 0.1982,
+      "step": 42000
+    },
+    {
+      "epoch": 0.5527188503447913,
+      "eval_loss": 0.2143728882074356,
+      "eval_runtime": 4.4114,
+      "eval_samples_per_second": 3.627,
+      "eval_steps_per_second": 1.813,
+      "step": 42000
+    },
+    {
+      "epoch": 0.5529820497973365,
+      "grad_norm": 1.2486026287078857,
+      "learning_rate": 4.386512360926407e-06,
+      "loss": 0.1848,
+      "step": 42020
+    },
+    {
+      "epoch": 0.5532452492498816,
+      "grad_norm": 1.4581921100616455,
+      "learning_rate": 4.38228257520166e-06,
+      "loss": 0.1772,
+      "step": 42040
+    },
+    {
+      "epoch": 0.5535084487024267,
+      "grad_norm": 1.5077170133590698,
+      "learning_rate": 4.3780532383466335e-06,
+      "loss": 0.1781,
+      "step": 42060
+    },
+    {
+      "epoch": 0.5537716481549718,
+      "grad_norm": 1.5865238904953003,
+      "learning_rate": 4.3738243534346085e-06,
+      "loss": 0.1812,
+      "step": 42080
+    },
+    {
+      "epoch": 0.554034847607517,
+      "grad_norm": 1.7937428951263428,
+      "learning_rate": 4.369595923538544e-06,
+      "loss": 0.1719,
+      "step": 42100
+    },
+    {
+      "epoch": 0.5542980470600621,
+      "grad_norm": 1.3743053674697876,
+      "learning_rate": 4.365367951731063e-06,
+      "loss": 0.1887,
+      "step": 42120
+    },
+    {
+      "epoch": 0.5545612465126073,
+      "grad_norm": 1.7606548070907593,
+      "learning_rate": 4.361140441084457e-06,
+      "loss": 0.1797,
+      "step": 42140
+    },
+    {
+      "epoch": 0.5548244459651523,
+      "grad_norm": 1.0714051723480225,
+      "learning_rate": 4.3569133946706865e-06,
+      "loss": 0.1801,
+      "step": 42160
+    },
+    {
+      "epoch": 0.5550876454176975,
+      "grad_norm": 1.6777467727661133,
+      "learning_rate": 4.352686815561368e-06,
+      "loss": 0.1723,
+      "step": 42180
+    },
+    {
+      "epoch": 0.5553508448702427,
+      "grad_norm": 1.6526973247528076,
+      "learning_rate": 4.348460706827784e-06,
+      "loss": 0.1754,
+      "step": 42200
+    },
+    {
+      "epoch": 0.5556140443227878,
+      "grad_norm": 1.550150752067566,
+      "learning_rate": 4.3442350715408734e-06,
+      "loss": 0.1573,
+      "step": 42220
+    },
+    {
+      "epoch": 0.555877243775333,
+      "grad_norm": 1.4112643003463745,
+      "learning_rate": 4.340009912771227e-06,
+      "loss": 0.1775,
+      "step": 42240
+    },
+    {
+      "epoch": 0.556140443227878,
+      "grad_norm": 1.648233413696289,
+      "learning_rate": 4.335785233589099e-06,
+      "loss": 0.2081,
+      "step": 42260
+    },
+    {
+      "epoch": 0.5564036426804232,
+      "grad_norm": 1.4796791076660156,
+      "learning_rate": 4.331561037064385e-06,
+      "loss": 0.1677,
+      "step": 42280
+    },
+    {
+      "epoch": 0.5566668421329684,
+      "grad_norm": 1.6772164106369019,
+      "learning_rate": 4.3273373262666326e-06,
+      "loss": 0.1777,
+      "step": 42300
+    },
+    {
+      "epoch": 0.5569300415855135,
+      "grad_norm": 1.4415826797485352,
+      "learning_rate": 4.323114104265041e-06,
+      "loss": 0.1835,
+      "step": 42320
+    },
+    {
+      "epoch": 0.5571932410380587,
+      "grad_norm": 1.821205496788025,
+      "learning_rate": 4.3188913741284505e-06,
+      "loss": 0.17,
+      "step": 42340
+    },
+    {
+      "epoch": 0.5574564404906038,
+      "grad_norm": 1.8219892978668213,
+      "learning_rate": 4.3146691389253446e-06,
+      "loss": 0.184,
+      "step": 42360
+    },
+    {
+      "epoch": 0.5577196399431489,
+      "grad_norm": 1.2814031839370728,
+      "learning_rate": 4.310447401723846e-06,
+      "loss": 0.1941,
+      "step": 42380
+    },
+    {
+      "epoch": 0.557982839395694,
+      "grad_norm": 2.0285565853118896,
+      "learning_rate": 4.306226165591717e-06,
+      "loss": 0.1966,
+      "step": 42400
+    },
+    {
+      "epoch": 0.5582460388482392,
+      "grad_norm": 1.5354300737380981,
+      "learning_rate": 4.302005433596357e-06,
+      "loss": 0.1717,
+      "step": 42420
+    },
+    {
+      "epoch": 0.5585092383007844,
+      "grad_norm": 1.3112965822219849,
+      "learning_rate": 4.297785208804794e-06,
+      "loss": 0.1918,
+      "step": 42440
+    },
+    {
+      "epoch": 0.5587724377533295,
+      "grad_norm": 1.8369892835617065,
+      "learning_rate": 4.293565494283696e-06,
+      "loss": 0.1763,
+      "step": 42460
+    },
+    {
+      "epoch": 0.5590356372058746,
+      "grad_norm": 1.6991682052612305,
+      "learning_rate": 4.28934629309935e-06,
+      "loss": 0.1654,
+      "step": 42480
+    },
+    {
+      "epoch": 0.5592988366584197,
+      "grad_norm": 1.2617453336715698,
+      "learning_rate": 4.285127608317677e-06,
+      "loss": 0.2029,
+      "step": 42500
+    },
+    {
+      "epoch": 0.5592988366584197,
+      "eval_loss": 0.2137726992368698,
+      "eval_runtime": 4.2906,
+      "eval_samples_per_second": 3.729,
+      "eval_steps_per_second": 1.865,
+      "step": 42500
+    },
+    {
+      "epoch": 0.5595620361109649,
+      "grad_norm": 1.6545333862304688,
+      "learning_rate": 4.280909443004223e-06,
+      "loss": 0.1926,
+      "step": 42520
+    },
+    {
+      "epoch": 0.55982523556351,
+      "grad_norm": 1.4431040287017822,
+      "learning_rate": 4.276691800224151e-06,
+      "loss": 0.1747,
+      "step": 42540
+    },
+    {
+      "epoch": 0.5600884350160552,
+      "grad_norm": 1.7165035009384155,
+      "learning_rate": 4.272474683042249e-06,
+      "loss": 0.1906,
+      "step": 42560
+    },
+    {
+      "epoch": 0.5603516344686003,
+      "grad_norm": 1.1916258335113525,
+      "learning_rate": 4.268258094522923e-06,
+      "loss": 0.1804,
+      "step": 42580
+    },
+    {
+      "epoch": 0.5606148339211454,
+      "grad_norm": 1.5146859884262085,
+      "learning_rate": 4.264042037730191e-06,
+      "loss": 0.1763,
+      "step": 42600
+    },
+    {
+      "epoch": 0.5608780333736906,
+      "grad_norm": 1.8203917741775513,
+      "learning_rate": 4.259826515727689e-06,
+      "loss": 0.1824,
+      "step": 42620
+    },
+    {
+      "epoch": 0.5611412328262357,
+      "grad_norm": 1.4621174335479736,
+      "learning_rate": 4.255611531578665e-06,
+      "loss": 0.187,
+      "step": 42640
+    },
+    {
+      "epoch": 0.5614044322787809,
+      "grad_norm": 1.3350205421447754,
+      "learning_rate": 4.2513970883459665e-06,
+      "loss": 0.18,
+      "step": 42660
+    },
+    {
+      "epoch": 0.561667631731326,
+      "grad_norm": 1.2826396226882935,
+      "learning_rate": 4.247183189092064e-06,
+      "loss": 0.1791,
+      "step": 42680
+    },
+    {
+      "epoch": 0.5619308311838711,
+      "grad_norm": 1.4072140455245972,
+      "learning_rate": 4.242969836879016e-06,
+      "loss": 0.1674,
+      "step": 42700
+    },
+    {
+      "epoch": 0.5621940306364163,
+      "grad_norm": 1.4222897291183472,
+      "learning_rate": 4.238757034768497e-06,
+      "loss": 0.2001,
+      "step": 42720
+    },
+    {
+      "epoch": 0.5624572300889614,
+      "grad_norm": 1.0692731142044067,
+      "learning_rate": 4.2345447858217744e-06,
+      "loss": 0.173,
+      "step": 42740
+    },
+    {
+      "epoch": 0.5627204295415066,
+      "grad_norm": 1.8840718269348145,
+      "learning_rate": 4.230333093099713e-06,
+      "loss": 0.1921,
+      "step": 42760
+    },
+    {
+      "epoch": 0.5629836289940517,
+      "grad_norm": 1.5556292533874512,
+      "learning_rate": 4.22612195966278e-06,
+      "loss": 0.1688,
+      "step": 42780
+    },
+    {
+      "epoch": 0.5632468284465968,
+      "grad_norm": 1.8242406845092773,
+      "learning_rate": 4.2219113885710285e-06,
+      "loss": 0.1821,
+      "step": 42800
+    },
+    {
+      "epoch": 0.563510027899142,
+      "grad_norm": 1.443302035331726,
+      "learning_rate": 4.217701382884109e-06,
+      "loss": 0.1816,
+      "step": 42820
+    },
+    {
+      "epoch": 0.5637732273516871,
+      "grad_norm": 1.328647255897522,
+      "learning_rate": 4.213491945661259e-06,
+      "loss": 0.1811,
+      "step": 42840
+    },
+    {
+      "epoch": 0.5640364268042323,
+      "grad_norm": 1.7687879800796509,
+      "learning_rate": 4.2092830799612975e-06,
+      "loss": 0.1619,
+      "step": 42860
+    },
+    {
+      "epoch": 0.5642996262567774,
+      "grad_norm": 1.3984055519104004,
+      "learning_rate": 4.205074788842639e-06,
+      "loss": 0.1619,
+      "step": 42880
+    },
+    {
+      "epoch": 0.5645628257093225,
+      "grad_norm": 1.9073807001113892,
+      "learning_rate": 4.200867075363272e-06,
+      "loss": 0.1904,
+      "step": 42900
+    },
+    {
+      "epoch": 0.5648260251618676,
+      "grad_norm": 1.3337582349777222,
+      "learning_rate": 4.196659942580768e-06,
+      "loss": 0.1853,
+      "step": 42920
+    },
+    {
+      "epoch": 0.5650892246144128,
+      "grad_norm": 1.4582470655441284,
+      "learning_rate": 4.192453393552277e-06,
+      "loss": 0.1965,
+      "step": 42940
+    },
+    {
+      "epoch": 0.565352424066958,
+      "grad_norm": 1.9661765098571777,
+      "learning_rate": 4.1882474313345244e-06,
+      "loss": 0.1964,
+      "step": 42960
+    },
+    {
+      "epoch": 0.5656156235195031,
+      "grad_norm": 1.288757562637329,
+      "learning_rate": 4.184042058983809e-06,
+      "loss": 0.1789,
+      "step": 42980
+    },
+    {
+      "epoch": 0.5658788229720482,
+      "grad_norm": 1.5039323568344116,
+      "learning_rate": 4.179837279556001e-06,
+      "loss": 0.1885,
+      "step": 43000
+    },
+    {
+      "epoch": 0.5658788229720482,
+      "eval_loss": 0.21332861483097076,
+      "eval_runtime": 4.2959,
+      "eval_samples_per_second": 3.724,
+      "eval_steps_per_second": 1.862,
+      "step": 43000
+    },
+    {
+      "epoch": 0.5661420224245933,
+      "grad_norm": 1.028756022453308,
+      "learning_rate": 4.175633096106539e-06,
+      "loss": 0.1839,
+      "step": 43020
+    },
+    {
+      "epoch": 0.5664052218771385,
+      "grad_norm": 1.5512914657592773,
+      "learning_rate": 4.171429511690431e-06,
+      "loss": 0.1734,
+      "step": 43040
+    },
+    {
+      "epoch": 0.5666684213296836,
+      "grad_norm": 1.1548620462417603,
+      "learning_rate": 4.167226529362247e-06,
+      "loss": 0.1786,
+      "step": 43060
+    },
+    {
+      "epoch": 0.5669316207822288,
+      "grad_norm": 1.3279780149459839,
+      "learning_rate": 4.163024152176122e-06,
+      "loss": 0.1711,
+      "step": 43080
+    },
+    {
+      "epoch": 0.567194820234774,
+      "grad_norm": 1.5263471603393555,
+      "learning_rate": 4.158822383185747e-06,
+      "loss": 0.2004,
+      "step": 43100
+    },
+    {
+      "epoch": 0.567458019687319,
+      "grad_norm": 1.6705269813537598,
+      "learning_rate": 4.154621225444374e-06,
+      "loss": 0.1634,
+      "step": 43120
+    },
+    {
+      "epoch": 0.5677212191398642,
+      "grad_norm": 2.0602221488952637,
+      "learning_rate": 4.150420682004814e-06,
+      "loss": 0.1874,
+      "step": 43140
+    },
+    {
+      "epoch": 0.5679844185924093,
+      "grad_norm": 2.346134901046753,
+      "learning_rate": 4.146220755919426e-06,
+      "loss": 0.1605,
+      "step": 43160
+    },
+    {
+      "epoch": 0.5682476180449545,
+      "grad_norm": 1.5779691934585571,
+      "learning_rate": 4.14202145024012e-06,
+      "loss": 0.1814,
+      "step": 43180
+    },
+    {
+      "epoch": 0.5685108174974997,
+      "grad_norm": 1.4583499431610107,
+      "learning_rate": 4.1378227680183645e-06,
+      "loss": 0.1843,
+      "step": 43200
+    },
+    {
+      "epoch": 0.5687740169500447,
+      "grad_norm": 1.4545915126800537,
+      "learning_rate": 4.133624712305162e-06,
+      "loss": 0.1972,
+      "step": 43220
+    },
+    {
+      "epoch": 0.5690372164025899,
+      "grad_norm": 1.6105595827102661,
+      "learning_rate": 4.12942728615107e-06,
+      "loss": 0.1753,
+      "step": 43240
+    },
+    {
+      "epoch": 0.569300415855135,
+      "grad_norm": 1.1848297119140625,
+      "learning_rate": 4.125230492606184e-06,
+      "loss": 0.1982,
+      "step": 43260
+    },
+    {
+      "epoch": 0.5695636153076802,
+      "grad_norm": 1.4252827167510986,
+      "learning_rate": 4.121034334720137e-06,
+      "loss": 0.1771,
+      "step": 43280
+    },
+    {
+      "epoch": 0.5698268147602253,
+      "grad_norm": 1.8780242204666138,
+      "learning_rate": 4.116838815542109e-06,
+      "loss": 0.1782,
+      "step": 43300
+    },
+    {
+      "epoch": 0.5700900142127704,
+      "grad_norm": 1.934571623802185,
+      "learning_rate": 4.112643938120806e-06,
+      "loss": 0.1748,
+      "step": 43320
+    },
+    {
+      "epoch": 0.5703532136653156,
+      "grad_norm": 1.434390664100647,
+      "learning_rate": 4.108449705504473e-06,
+      "loss": 0.1805,
+      "step": 43340
+    },
+    {
+      "epoch": 0.5706164131178607,
+      "grad_norm": 1.791860818862915,
+      "learning_rate": 4.104256120740887e-06,
+      "loss": 0.1868,
+      "step": 43360
+    },
+    {
+      "epoch": 0.5708796125704059,
+      "grad_norm": 1.6886893510818481,
+      "learning_rate": 4.10006318687735e-06,
+      "loss": 0.1885,
+      "step": 43380
+    },
+    {
+      "epoch": 0.571142812022951,
+      "grad_norm": 2.0041117668151855,
+      "learning_rate": 4.095870906960697e-06,
+      "loss": 0.1832,
+      "step": 43400
+    },
+    {
+      "epoch": 0.5714060114754961,
+      "grad_norm": 1.5318540334701538,
+      "learning_rate": 4.091679284037284e-06,
+      "loss": 0.1912,
+      "step": 43420
+    },
+    {
+      "epoch": 0.5716692109280412,
+      "grad_norm": 1.633003830909729,
+      "learning_rate": 4.087488321152986e-06,
+      "loss": 0.1889,
+      "step": 43440
+    },
+    {
+      "epoch": 0.5719324103805864,
+      "grad_norm": 1.93812096118927,
+      "learning_rate": 4.0832980213532075e-06,
+      "loss": 0.1848,
+      "step": 43460
+    },
+    {
+      "epoch": 0.5721956098331316,
+      "grad_norm": 1.719903826713562,
+      "learning_rate": 4.0791083876828615e-06,
+      "loss": 0.1709,
+      "step": 43480
+    },
+    {
+      "epoch": 0.5724588092856767,
+      "grad_norm": 1.9009802341461182,
+      "learning_rate": 4.0749194231863875e-06,
+      "loss": 0.1862,
+      "step": 43500
+    },
+    {
+      "epoch": 0.5724588092856767,
+      "eval_loss": 0.20632830262184143,
+      "eval_runtime": 4.389,
+      "eval_samples_per_second": 3.645,
+      "eval_steps_per_second": 1.823,
+      "step": 43500
+    },
+    {
+      "epoch": 0.5727220087382219,
+      "grad_norm": 1.2185953855514526,
+      "learning_rate": 4.070731130907729e-06,
+      "loss": 0.159,
+      "step": 43520
+    },
+    {
+      "epoch": 0.5729852081907669,
+      "grad_norm": 1.328132152557373,
+      "learning_rate": 4.066543513890344e-06,
+      "loss": 0.1853,
+      "step": 43540
+    },
+    {
+      "epoch": 0.5732484076433121,
+      "grad_norm": 1.5405911207199097,
+      "learning_rate": 4.0623565751772035e-06,
+      "loss": 0.1653,
+      "step": 43560
+    },
+    {
+      "epoch": 0.5735116070958572,
+      "grad_norm": 1.4149196147918701,
+      "learning_rate": 4.05817031781078e-06,
+      "loss": 0.1813,
+      "step": 43580
+    },
+    {
+      "epoch": 0.5737748065484024,
+      "grad_norm": 1.1772006750106812,
+      "learning_rate": 4.053984744833056e-06,
+      "loss": 0.188,
+      "step": 43600
+    },
+    {
+      "epoch": 0.5740380060009476,
+      "grad_norm": 1.561081886291504,
+      "learning_rate": 4.049799859285514e-06,
+      "loss": 0.1726,
+      "step": 43620
+    },
+    {
+      "epoch": 0.5743012054534926,
+      "grad_norm": 1.242539882659912,
+      "learning_rate": 4.045615664209136e-06,
+      "loss": 0.1963,
+      "step": 43640
+    },
+    {
+      "epoch": 0.5745644049060378,
+      "grad_norm": 1.411284327507019,
+      "learning_rate": 4.041432162644404e-06,
+      "loss": 0.1898,
+      "step": 43660
+    },
+    {
+      "epoch": 0.5748276043585829,
+      "grad_norm": 1.6267142295837402,
+      "learning_rate": 4.037249357631299e-06,
+      "loss": 0.1778,
+      "step": 43680
+    },
+    {
+      "epoch": 0.5750908038111281,
+      "grad_norm": 1.4784654378890991,
+      "learning_rate": 4.033067252209286e-06,
+      "loss": 0.19,
+      "step": 43700
+    },
+    {
+      "epoch": 0.5753540032636733,
+      "grad_norm": 1.6932753324508667,
+      "learning_rate": 4.0288858494173334e-06,
+      "loss": 0.1833,
+      "step": 43720
+    },
+    {
+      "epoch": 0.5756172027162183,
+      "grad_norm": 1.4818893671035767,
+      "learning_rate": 4.0247051522938915e-06,
+      "loss": 0.1762,
+      "step": 43740
+    },
+    {
+      "epoch": 0.5758804021687635,
+      "grad_norm": 1.8054322004318237,
+      "learning_rate": 4.020525163876902e-06,
+      "loss": 0.1885,
+      "step": 43760
+    },
+    {
+      "epoch": 0.5761436016213086,
+      "grad_norm": 1.836157202720642,
+      "learning_rate": 4.016345887203788e-06,
+      "loss": 0.1885,
+      "step": 43780
+    },
+    {
+      "epoch": 0.5764068010738538,
+      "grad_norm": 1.5319446325302124,
+      "learning_rate": 4.012167325311456e-06,
+      "loss": 0.1711,
+      "step": 43800
+    },
+    {
+      "epoch": 0.5766700005263989,
+      "grad_norm": 1.2254046201705933,
+      "learning_rate": 4.007989481236297e-06,
+      "loss": 0.181,
+      "step": 43820
+    },
+    {
+      "epoch": 0.576933199978944,
+      "grad_norm": 1.5358214378356934,
+      "learning_rate": 4.003812358014175e-06,
+      "loss": 0.1687,
+      "step": 43840
+    },
+    {
+      "epoch": 0.5771963994314891,
+      "grad_norm": 1.1001297235488892,
+      "learning_rate": 3.999635958680436e-06,
+      "loss": 0.1586,
+      "step": 43860
+    },
+    {
+      "epoch": 0.5774595988840343,
+      "grad_norm": 1.8449863195419312,
+      "learning_rate": 3.995460286269895e-06,
+      "loss": 0.191,
+      "step": 43880
+    },
+    {
+      "epoch": 0.5777227983365795,
+      "grad_norm": 1.2807457447052002,
+      "learning_rate": 3.991285343816837e-06,
+      "loss": 0.201,
+      "step": 43900
+    },
+    {
+      "epoch": 0.5779859977891246,
+      "grad_norm": 1.0096651315689087,
+      "learning_rate": 3.987111134355027e-06,
+      "loss": 0.1699,
+      "step": 43920
+    },
+    {
+      "epoch": 0.5782491972416698,
+      "grad_norm": 1.8587656021118164,
+      "learning_rate": 3.982937660917687e-06,
+      "loss": 0.1941,
+      "step": 43940
+    },
+    {
+      "epoch": 0.5785123966942148,
+      "grad_norm": 1.6979950666427612,
+      "learning_rate": 3.978764926537505e-06,
+      "loss": 0.1909,
+      "step": 43960
+    },
+    {
+      "epoch": 0.57877559614676,
+      "grad_norm": 1.321791648864746,
+      "learning_rate": 3.974592934246639e-06,
+      "loss": 0.176,
+      "step": 43980
+    },
+    {
+      "epoch": 0.5790387955993052,
+      "grad_norm": 1.410238265991211,
+      "learning_rate": 3.9704216870767e-06,
+      "loss": 0.1646,
+      "step": 44000
+    },
+    {
+      "epoch": 0.5790387955993052,
+      "eval_loss": 0.208803191781044,
+      "eval_runtime": 3.6644,
+      "eval_samples_per_second": 4.366,
+      "eval_steps_per_second": 2.183,
+      "step": 44000
+    },
+    {
+      "epoch": 0.5793019950518503,
+      "grad_norm": 1.826768159866333,
+      "learning_rate": 3.966251188058764e-06,
+      "loss": 0.2022,
+      "step": 44020
+    },
+    {
+      "epoch": 0.5795651945043955,
+      "grad_norm": 1.7175239324569702,
+      "learning_rate": 3.9620814402233585e-06,
+      "loss": 0.1719,
+      "step": 44040
+    },
+    {
+      "epoch": 0.5798283939569405,
+      "grad_norm": 1.4297294616699219,
+      "learning_rate": 3.957912446600465e-06,
+      "loss": 0.2021,
+      "step": 44060
+    },
+    {
+      "epoch": 0.5800915934094857,
+      "grad_norm": 1.4260920286178589,
+      "learning_rate": 3.953744210219525e-06,
+      "loss": 0.1902,
+      "step": 44080
+    },
+    {
+      "epoch": 0.5803547928620308,
+      "grad_norm": 1.5720634460449219,
+      "learning_rate": 3.9495767341094146e-06,
+      "loss": 0.1834,
+      "step": 44100
+    },
+    {
+      "epoch": 0.580617992314576,
+      "grad_norm": 1.754641056060791,
+      "learning_rate": 3.945410021298475e-06,
+      "loss": 0.1668,
+      "step": 44120
+    },
+    {
+      "epoch": 0.5808811917671212,
+      "grad_norm": 1.4831589460372925,
+      "learning_rate": 3.941244074814479e-06,
+      "loss": 0.1729,
+      "step": 44140
+    },
+    {
+      "epoch": 0.5811443912196662,
+      "grad_norm": 1.4108664989471436,
+      "learning_rate": 3.937078897684648e-06,
+      "loss": 0.1727,
+      "step": 44160
+    },
+    {
+      "epoch": 0.5814075906722114,
+      "grad_norm": 1.832789659500122,
+      "learning_rate": 3.9329144929356475e-06,
+      "loss": 0.1777,
+      "step": 44180
+    },
+    {
+      "epoch": 0.5816707901247565,
+      "grad_norm": 1.6369285583496094,
+      "learning_rate": 3.928750863593574e-06,
+      "loss": 0.1936,
+      "step": 44200
+    },
+    {
+      "epoch": 0.5819339895773017,
+      "grad_norm": 1.6507072448730469,
+      "learning_rate": 3.924588012683969e-06,
+      "loss": 0.1831,
+      "step": 44220
+    },
+    {
+      "epoch": 0.5821971890298469,
+      "grad_norm": 1.4222640991210938,
+      "learning_rate": 3.9204259432318e-06,
+      "loss": 0.1965,
+      "step": 44240
+    },
+    {
+      "epoch": 0.582460388482392,
+      "grad_norm": 1.8923921585083008,
+      "learning_rate": 3.9162646582614736e-06,
+      "loss": 0.1835,
+      "step": 44260
+    },
+    {
+      "epoch": 0.5827235879349371,
+      "grad_norm": 1.8329373598098755,
+      "learning_rate": 3.9121041607968226e-06,
+      "loss": 0.2006,
+      "step": 44280
+    },
+    {
+      "epoch": 0.5829867873874822,
+      "grad_norm": 1.5261905193328857,
+      "learning_rate": 3.907944453861111e-06,
+      "loss": 0.1786,
+      "step": 44300
+    },
+    {
+      "epoch": 0.5832499868400274,
+      "grad_norm": 1.33411705493927,
+      "learning_rate": 3.90378554047702e-06,
+      "loss": 0.1864,
+      "step": 44320
+    },
+    {
+      "epoch": 0.5835131862925725,
+      "grad_norm": 1.872057318687439,
+      "learning_rate": 3.8996274236666666e-06,
+      "loss": 0.1872,
+      "step": 44340
+    },
+    {
+      "epoch": 0.5837763857451177,
+      "grad_norm": 1.285286784172058,
+      "learning_rate": 3.895470106451577e-06,
+      "loss": 0.1757,
+      "step": 44360
+    },
+    {
+      "epoch": 0.5840395851976627,
+      "grad_norm": 1.4806030988693237,
+      "learning_rate": 3.891313591852706e-06,
+      "loss": 0.1792,
+      "step": 44380
+    },
+    {
+      "epoch": 0.5843027846502079,
+      "grad_norm": 1.9090381860733032,
+      "learning_rate": 3.887157882890419e-06,
+      "loss": 0.1929,
+      "step": 44400
+    },
+    {
+      "epoch": 0.5845659841027531,
+      "grad_norm": 1.2179738283157349,
+      "learning_rate": 3.883002982584497e-06,
+      "loss": 0.1928,
+      "step": 44420
+    },
+    {
+      "epoch": 0.5848291835552982,
+      "grad_norm": 1.2604807615280151,
+      "learning_rate": 3.878848893954136e-06,
+      "loss": 0.2062,
+      "step": 44440
+    },
+    {
+      "epoch": 0.5850923830078434,
+      "grad_norm": 1.5589056015014648,
+      "learning_rate": 3.8746956200179395e-06,
+      "loss": 0.172,
+      "step": 44460
+    },
+    {
+      "epoch": 0.5853555824603884,
+      "grad_norm": 1.3426591157913208,
+      "learning_rate": 3.870543163793922e-06,
+      "loss": 0.1966,
+      "step": 44480
+    },
+    {
+      "epoch": 0.5856187819129336,
+      "grad_norm": 1.7154895067214966,
+      "learning_rate": 3.8663915282995e-06,
+      "loss": 0.1863,
+      "step": 44500
+    },
+    {
+      "epoch": 0.5856187819129336,
+      "eval_loss": 0.21264442801475525,
+      "eval_runtime": 4.4515,
+      "eval_samples_per_second": 3.594,
+      "eval_steps_per_second": 1.797,
+      "step": 44500
+    },
+    {
+      "epoch": 0.5858819813654788,
+      "grad_norm": 2.0353856086730957,
+      "learning_rate": 3.862240716551494e-06,
+      "loss": 0.1798,
+      "step": 44520
+    },
+    {
+      "epoch": 0.5861451808180239,
+      "grad_norm": 1.1277720928192139,
+      "learning_rate": 3.858090731566131e-06,
+      "loss": 0.1894,
+      "step": 44540
+    },
+    {
+      "epoch": 0.5864083802705691,
+      "grad_norm": 1.2137243747711182,
+      "learning_rate": 3.853941576359031e-06,
+      "loss": 0.174,
+      "step": 44560
+    },
+    {
+      "epoch": 0.5866715797231141,
+      "grad_norm": 1.678031325340271,
+      "learning_rate": 3.8497932539452125e-06,
+      "loss": 0.1629,
+      "step": 44580
+    },
+    {
+      "epoch": 0.5869347791756593,
+      "grad_norm": 1.6398566961288452,
+      "learning_rate": 3.845645767339092e-06,
+      "loss": 0.1819,
+      "step": 44600
+    },
+    {
+      "epoch": 0.5871979786282044,
+      "grad_norm": 1.235878586769104,
+      "learning_rate": 3.841499119554476e-06,
+      "loss": 0.191,
+      "step": 44620
+    },
+    {
+      "epoch": 0.5874611780807496,
+      "grad_norm": 1.6539210081100464,
+      "learning_rate": 3.837353313604562e-06,
+      "loss": 0.1748,
+      "step": 44640
+    },
+    {
+      "epoch": 0.5877243775332948,
+      "grad_norm": 1.3157024383544922,
+      "learning_rate": 3.833208352501937e-06,
+      "loss": 0.1736,
+      "step": 44660
+    },
+    {
+      "epoch": 0.5879875769858399,
+      "grad_norm": 1.427213430404663,
+      "learning_rate": 3.82906423925857e-06,
+      "loss": 0.1674,
+      "step": 44680
+    },
+    {
+      "epoch": 0.588250776438385,
+      "grad_norm": 1.4954136610031128,
+      "learning_rate": 3.824920976885821e-06,
+      "loss": 0.1921,
+      "step": 44700
+    },
+    {
+      "epoch": 0.5885139758909301,
+      "grad_norm": 1.4666842222213745,
+      "learning_rate": 3.820778568394423e-06,
+      "loss": 0.1715,
+      "step": 44720
+    },
+    {
+      "epoch": 0.5887771753434753,
+      "grad_norm": 1.614347219467163,
+      "learning_rate": 3.816637016794496e-06,
+      "loss": 0.1959,
+      "step": 44740
+    },
+    {
+      "epoch": 0.5890403747960204,
+      "grad_norm": 1.4376959800720215,
+      "learning_rate": 3.8124963250955348e-06,
+      "loss": 0.1792,
+      "step": 44760
+    },
+    {
+      "epoch": 0.5893035742485656,
+      "grad_norm": 1.7915065288543701,
+      "learning_rate": 3.8083564963064053e-06,
+      "loss": 0.1918,
+      "step": 44780
+    },
+    {
+      "epoch": 0.5895667737011107,
+      "grad_norm": 1.5536233186721802,
+      "learning_rate": 3.8042175334353537e-06,
+      "loss": 0.1812,
+      "step": 44800
+    },
+    {
+      "epoch": 0.5898299731536558,
+      "grad_norm": 1.385319471359253,
+      "learning_rate": 3.8000794394899915e-06,
+      "loss": 0.1736,
+      "step": 44820
+    },
+    {
+      "epoch": 0.590093172606201,
+      "grad_norm": 1.5073901414871216,
+      "learning_rate": 3.7959422174772998e-06,
+      "loss": 0.1864,
+      "step": 44840
+    },
+    {
+      "epoch": 0.5903563720587461,
+      "grad_norm": 1.9643348455429077,
+      "learning_rate": 3.7918058704036287e-06,
+      "loss": 0.182,
+      "step": 44860
+    },
+    {
+      "epoch": 0.5906195715112913,
+      "grad_norm": 1.6934657096862793,
+      "learning_rate": 3.7876704012746884e-06,
+      "loss": 0.1755,
+      "step": 44880
+    },
+    {
+      "epoch": 0.5908827709638363,
+      "grad_norm": 1.1022480726242065,
+      "learning_rate": 3.7835358130955564e-06,
+      "loss": 0.1711,
+      "step": 44900
+    },
+    {
+      "epoch": 0.5911459704163815,
+      "grad_norm": 1.6657503843307495,
+      "learning_rate": 3.7794021088706662e-06,
+      "loss": 0.1861,
+      "step": 44920
+    },
+    {
+      "epoch": 0.5914091698689267,
+      "grad_norm": 1.6407934427261353,
+      "learning_rate": 3.775269291603805e-06,
+      "loss": 0.1781,
+      "step": 44940
+    },
+    {
+      "epoch": 0.5916723693214718,
+      "grad_norm": 1.3959637880325317,
+      "learning_rate": 3.7711373642981287e-06,
+      "loss": 0.1676,
+      "step": 44960
+    },
+    {
+      "epoch": 0.591935568774017,
+      "grad_norm": 1.4866254329681396,
+      "learning_rate": 3.7670063299561305e-06,
+      "loss": 0.1859,
+      "step": 44980
+    },
+    {
+      "epoch": 0.592198768226562,
+      "grad_norm": 1.7329636812210083,
+      "learning_rate": 3.7628761915796667e-06,
+      "loss": 0.1883,
+      "step": 45000
+    },
+    {
+      "epoch": 0.592198768226562,
+      "eval_loss": 0.21048158407211304,
+      "eval_runtime": 3.1373,
+      "eval_samples_per_second": 5.1,
+      "eval_steps_per_second": 2.55,
+      "step": 45000
+    },
+    {
+      "epoch": 0.5924619676791072,
+      "grad_norm": 1.4540742635726929,
+      "learning_rate": 3.758746952169937e-06,
+      "loss": 0.1768,
+      "step": 45020
+    },
+    {
+      "epoch": 0.5927251671316524,
+      "grad_norm": 1.4844151735305786,
+      "learning_rate": 3.7546186147274878e-06,
+      "loss": 0.1818,
+      "step": 45040
+    },
+    {
+      "epoch": 0.5929883665841975,
+      "grad_norm": 1.8462321758270264,
+      "learning_rate": 3.750491182252213e-06,
+      "loss": 0.1744,
+      "step": 45060
+    },
+    {
+      "epoch": 0.5932515660367427,
+      "grad_norm": 2.079591751098633,
+      "learning_rate": 3.746364657743347e-06,
+      "loss": 0.1767,
+      "step": 45080
+    },
+    {
+      "epoch": 0.5935147654892878,
+      "grad_norm": 2.4083993434906006,
+      "learning_rate": 3.7422390441994615e-06,
+      "loss": 0.1761,
+      "step": 45100
+    },
+    {
+      "epoch": 0.5937779649418329,
+      "grad_norm": 1.5558803081512451,
+      "learning_rate": 3.7381143446184753e-06,
+      "loss": 0.1785,
+      "step": 45120
+    },
+    {
+      "epoch": 0.594041164394378,
+      "grad_norm": 1.787089228630066,
+      "learning_rate": 3.7339905619976296e-06,
+      "loss": 0.1935,
+      "step": 45140
+    },
+    {
+      "epoch": 0.5943043638469232,
+      "grad_norm": 2.0632967948913574,
+      "learning_rate": 3.729867699333515e-06,
+      "loss": 0.1855,
+      "step": 45160
+    },
+    {
+      "epoch": 0.5945675632994684,
+      "grad_norm": 1.6688904762268066,
+      "learning_rate": 3.7257457596220392e-06,
+      "loss": 0.1635,
+      "step": 45180
+    },
+    {
+      "epoch": 0.5948307627520135,
+      "grad_norm": 1.4397289752960205,
+      "learning_rate": 3.7216247458584467e-06,
+      "loss": 0.1778,
+      "step": 45200
+    },
+    {
+      "epoch": 0.5950939622045586,
+      "grad_norm": 1.6527363061904907,
+      "learning_rate": 3.7175046610373095e-06,
+      "loss": 0.1868,
+      "step": 45220
+    },
+    {
+      "epoch": 0.5953571616571037,
+      "grad_norm": 1.7847778797149658,
+      "learning_rate": 3.7133855081525212e-06,
+      "loss": 0.1928,
+      "step": 45240
+    },
+    {
+      "epoch": 0.5956203611096489,
+      "grad_norm": 1.473523736000061,
+      "learning_rate": 3.7092672901973014e-06,
+      "loss": 0.1844,
+      "step": 45260
+    },
+    {
+      "epoch": 0.595883560562194,
+      "grad_norm": 1.4380725622177124,
+      "learning_rate": 3.705150010164188e-06,
+      "loss": 0.204,
+      "step": 45280
+    },
+    {
+      "epoch": 0.5961467600147392,
+      "grad_norm": 1.5294369459152222,
+      "learning_rate": 3.7010336710450377e-06,
+      "loss": 0.1906,
+      "step": 45300
+    },
+    {
+      "epoch": 0.5964099594672843,
+      "grad_norm": 2.1970791816711426,
+      "learning_rate": 3.6969182758310258e-06,
+      "loss": 0.182,
+      "step": 45320
+    },
+    {
+      "epoch": 0.5966731589198294,
+      "grad_norm": 1.4584088325500488,
+      "learning_rate": 3.69280382751264e-06,
+      "loss": 0.1959,
+      "step": 45340
+    },
+    {
+      "epoch": 0.5969363583723746,
+      "grad_norm": 1.2210859060287476,
+      "learning_rate": 3.688690329079676e-06,
+      "loss": 0.1926,
+      "step": 45360
+    },
+    {
+      "epoch": 0.5971995578249197,
+      "grad_norm": 1.5783464908599854,
+      "learning_rate": 3.6845777835212465e-06,
+      "loss": 0.1634,
+      "step": 45380
+    },
+    {
+      "epoch": 0.5974627572774649,
+      "grad_norm": 1.2974082231521606,
+      "learning_rate": 3.6804661938257667e-06,
+      "loss": 0.1771,
+      "step": 45400
+    },
+    {
+      "epoch": 0.59772595673001,
+      "grad_norm": 1.863672137260437,
+      "learning_rate": 3.676355562980961e-06,
+      "loss": 0.1912,
+      "step": 45420
+    },
+    {
+      "epoch": 0.5979891561825551,
+      "grad_norm": 1.4885729551315308,
+      "learning_rate": 3.672245893973853e-06,
+      "loss": 0.2027,
+      "step": 45440
+    },
+    {
+      "epoch": 0.5982523556351003,
+      "grad_norm": 1.5184351205825806,
+      "learning_rate": 3.668137189790769e-06,
+      "loss": 0.1846,
+      "step": 45460
+    },
+    {
+      "epoch": 0.5985155550876454,
+      "grad_norm": 1.4432047605514526,
+      "learning_rate": 3.664029453417337e-06,
+      "loss": 0.1873,
+      "step": 45480
+    },
+    {
+      "epoch": 0.5987787545401906,
+      "grad_norm": 1.306997299194336,
+      "learning_rate": 3.6599226878384764e-06,
+      "loss": 0.1908,
+      "step": 45500
+    },
+    {
+      "epoch": 0.5987787545401906,
+      "eval_loss": 0.2081170678138733,
+      "eval_runtime": 3.4662,
+      "eval_samples_per_second": 4.616,
+      "eval_steps_per_second": 2.308,
+      "step": 45500
+    },
+    {
+      "epoch": 0.5990419539927357,
+      "grad_norm": 1.4865301847457886,
+      "learning_rate": 3.6558168960384054e-06,
+      "loss": 0.2029,
+      "step": 45520
+    },
+    {
+      "epoch": 0.5993051534452808,
+      "grad_norm": 1.6525959968566895,
+      "learning_rate": 3.6517120810006346e-06,
+      "loss": 0.1841,
+      "step": 45540
+    },
+    {
+      "epoch": 0.599568352897826,
+      "grad_norm": 1.3946092128753662,
+      "learning_rate": 3.647608245707959e-06,
+      "loss": 0.1679,
+      "step": 45560
+    },
+    {
+      "epoch": 0.5998315523503711,
+      "grad_norm": 1.0273364782333374,
+      "learning_rate": 3.6435053931424713e-06,
+      "loss": 0.188,
+      "step": 45580
+    },
+    {
+      "epoch": 0.6000947518029163,
+      "grad_norm": 1.3998351097106934,
+      "learning_rate": 3.63940352628554e-06,
+      "loss": 0.1681,
+      "step": 45600
+    },
+    {
+      "epoch": 0.6003579512554614,
+      "grad_norm": 1.4155207872390747,
+      "learning_rate": 3.6353026481178274e-06,
+      "loss": 0.1822,
+      "step": 45620
+    },
+    {
+      "epoch": 0.6006211507080065,
+      "grad_norm": 1.2189795970916748,
+      "learning_rate": 3.631202761619268e-06,
+      "loss": 0.1852,
+      "step": 45640
+    },
+    {
+      "epoch": 0.6008843501605516,
+      "grad_norm": 1.7389336824417114,
+      "learning_rate": 3.627103869769081e-06,
+      "loss": 0.1803,
+      "step": 45660
+    },
+    {
+      "epoch": 0.6011475496130968,
+      "grad_norm": 1.8520710468292236,
+      "learning_rate": 3.6230059755457616e-06,
+      "loss": 0.175,
+      "step": 45680
+    },
+    {
+      "epoch": 0.601410749065642,
+      "grad_norm": 1.7885749340057373,
+      "learning_rate": 3.618909081927081e-06,
+      "loss": 0.1728,
+      "step": 45700
+    },
+    {
+      "epoch": 0.6016739485181871,
+      "grad_norm": 1.2058920860290527,
+      "learning_rate": 3.6148131918900815e-06,
+      "loss": 0.177,
+      "step": 45720
+    },
+    {
+      "epoch": 0.6019371479707322,
+      "grad_norm": 1.955143690109253,
+      "learning_rate": 3.6107183084110786e-06,
+      "loss": 0.1736,
+      "step": 45740
+    },
+    {
+      "epoch": 0.6022003474232773,
+      "grad_norm": 1.7161537408828735,
+      "learning_rate": 3.6066244344656515e-06,
+      "loss": 0.1924,
+      "step": 45760
+    },
+    {
+      "epoch": 0.6024635468758225,
+      "grad_norm": 1.3201004266738892,
+      "learning_rate": 3.6025315730286537e-06,
+      "loss": 0.1768,
+      "step": 45780
+    },
+    {
+      "epoch": 0.6027267463283676,
+      "grad_norm": 1.6143702268600464,
+      "learning_rate": 3.5984397270741954e-06,
+      "loss": 0.1918,
+      "step": 45800
+    },
+    {
+      "epoch": 0.6029899457809128,
+      "grad_norm": 1.6883364915847778,
+      "learning_rate": 3.5943488995756513e-06,
+      "loss": 0.188,
+      "step": 45820
+    },
+    {
+      "epoch": 0.603253145233458,
+      "grad_norm": 2.037290096282959,
+      "learning_rate": 3.5902590935056595e-06,
+      "loss": 0.1809,
+      "step": 45840
+    },
+    {
+      "epoch": 0.603516344686003,
+      "grad_norm": 1.6805565357208252,
+      "learning_rate": 3.58617031183611e-06,
+      "loss": 0.1828,
+      "step": 45860
+    },
+    {
+      "epoch": 0.6037795441385482,
+      "grad_norm": 1.618918538093567,
+      "learning_rate": 3.5820825575381545e-06,
+      "loss": 0.1797,
+      "step": 45880
+    },
+    {
+      "epoch": 0.6040427435910933,
+      "grad_norm": 1.4193322658538818,
+      "learning_rate": 3.5779958335821928e-06,
+      "loss": 0.1848,
+      "step": 45900
+    },
+    {
+      "epoch": 0.6043059430436385,
+      "grad_norm": 1.4136922359466553,
+      "learning_rate": 3.5739101429378784e-06,
+      "loss": 0.1864,
+      "step": 45920
+    },
+    {
+      "epoch": 0.6045691424961837,
+      "grad_norm": 1.2748030424118042,
+      "learning_rate": 3.5698254885741158e-06,
+      "loss": 0.1902,
+      "step": 45940
+    },
+    {
+      "epoch": 0.6048323419487287,
+      "grad_norm": 1.4001498222351074,
+      "learning_rate": 3.565741873459053e-06,
+      "loss": 0.1776,
+      "step": 45960
+    },
+    {
+      "epoch": 0.6050955414012739,
+      "grad_norm": 1.0041859149932861,
+      "learning_rate": 3.5616593005600818e-06,
+      "loss": 0.1821,
+      "step": 45980
+    },
+    {
+      "epoch": 0.605358740853819,
+      "grad_norm": 1.347687840461731,
+      "learning_rate": 3.5575777728438454e-06,
+      "loss": 0.1633,
+      "step": 46000
+    },
+    {
+      "epoch": 0.605358740853819,
+      "eval_loss": 0.217705637216568,
+      "eval_runtime": 4.2558,
+      "eval_samples_per_second": 3.76,
+      "eval_steps_per_second": 1.88,
+      "step": 46000
+    },
+    {
+      "epoch": 0.6056219403063642,
+      "grad_norm": 1.5138543844223022,
+      "learning_rate": 3.5534972932762147e-06,
+      "loss": 0.1764,
+      "step": 46020
+    },
+    {
+      "epoch": 0.6058851397589093,
+      "grad_norm": 1.6440802812576294,
+      "learning_rate": 3.54941786482231e-06,
+      "loss": 0.1952,
+      "step": 46040
+    },
+    {
+      "epoch": 0.6061483392114544,
+      "grad_norm": 1.6665791273117065,
+      "learning_rate": 3.545339490446481e-06,
+      "loss": 0.1779,
+      "step": 46060
+    },
+    {
+      "epoch": 0.6064115386639995,
+      "grad_norm": 1.6877797842025757,
+      "learning_rate": 3.5412621731123142e-06,
+      "loss": 0.1753,
+      "step": 46080
+    },
+    {
+      "epoch": 0.6066747381165447,
+      "grad_norm": 2.017784833908081,
+      "learning_rate": 3.5371859157826284e-06,
+      "loss": 0.1709,
+      "step": 46100
+    },
+    {
+      "epoch": 0.6069379375690899,
+      "grad_norm": 1.564407229423523,
+      "learning_rate": 3.5331107214194705e-06,
+      "loss": 0.1773,
+      "step": 46120
+    },
+    {
+      "epoch": 0.607201137021635,
+      "grad_norm": 1.36459219455719,
+      "learning_rate": 3.5290365929841165e-06,
+      "loss": 0.1863,
+      "step": 46140
+    },
+    {
+      "epoch": 0.6074643364741802,
+      "grad_norm": 1.645613193511963,
+      "learning_rate": 3.524963533437068e-06,
+      "loss": 0.1701,
+      "step": 46160
+    },
+    {
+      "epoch": 0.6077275359267252,
+      "grad_norm": 1.3956373929977417,
+      "learning_rate": 3.5208915457380445e-06,
+      "loss": 0.1772,
+      "step": 46180
+    },
+    {
+      "epoch": 0.6079907353792704,
+      "grad_norm": 1.78632652759552,
+      "learning_rate": 3.5168206328459977e-06,
+      "loss": 0.1651,
+      "step": 46200
+    },
+    {
+      "epoch": 0.6082539348318156,
+      "grad_norm": 1.497397780418396,
+      "learning_rate": 3.5127507977190867e-06,
+      "loss": 0.1891,
+      "step": 46220
+    },
+    {
+      "epoch": 0.6085171342843607,
+      "grad_norm": 1.1809332370758057,
+      "learning_rate": 3.5086820433146932e-06,
+      "loss": 0.1689,
+      "step": 46240
+    },
+    {
+      "epoch": 0.6087803337369059,
+      "grad_norm": 1.718700647354126,
+      "learning_rate": 3.5046143725894154e-06,
+      "loss": 0.1943,
+      "step": 46260
+    },
+    {
+      "epoch": 0.6090435331894509,
+      "grad_norm": 1.1331132650375366,
+      "learning_rate": 3.500547788499059e-06,
+      "loss": 0.1903,
+      "step": 46280
+    },
+    {
+      "epoch": 0.6093067326419961,
+      "grad_norm": 1.4409655332565308,
+      "learning_rate": 3.4964822939986437e-06,
+      "loss": 0.1696,
+      "step": 46300
+    },
+    {
+      "epoch": 0.6095699320945412,
+      "grad_norm": 1.4222794771194458,
+      "learning_rate": 3.492417892042397e-06,
+      "loss": 0.186,
+      "step": 46320
+    },
+    {
+      "epoch": 0.6098331315470864,
+      "grad_norm": 1.6676915884017944,
+      "learning_rate": 3.4883545855837487e-06,
+      "loss": 0.1698,
+      "step": 46340
+    },
+    {
+      "epoch": 0.6100963309996316,
+      "grad_norm": 1.6660655736923218,
+      "learning_rate": 3.48429237757534e-06,
+      "loss": 0.1613,
+      "step": 46360
+    },
+    {
+      "epoch": 0.6103595304521766,
+      "grad_norm": 1.3301000595092773,
+      "learning_rate": 3.480231270969005e-06,
+      "loss": 0.1827,
+      "step": 46380
+    },
+    {
+      "epoch": 0.6106227299047218,
+      "grad_norm": 1.5385206937789917,
+      "learning_rate": 3.4761712687157867e-06,
+      "loss": 0.1972,
+      "step": 46400
+    },
+    {
+      "epoch": 0.6108859293572669,
+      "grad_norm": 1.547641634941101,
+      "learning_rate": 3.4721123737659177e-06,
+      "loss": 0.172,
+      "step": 46420
+    },
+    {
+      "epoch": 0.6111491288098121,
+      "grad_norm": 1.5216346979141235,
+      "learning_rate": 3.4680545890688293e-06,
+      "loss": 0.1755,
+      "step": 46440
+    },
+    {
+      "epoch": 0.6114123282623573,
+      "grad_norm": 1.8222156763076782,
+      "learning_rate": 3.4639979175731476e-06,
+      "loss": 0.1755,
+      "step": 46460
+    },
+    {
+      "epoch": 0.6116755277149023,
+      "grad_norm": 1.709501028060913,
+      "learning_rate": 3.459942362226687e-06,
+      "loss": 0.1754,
+      "step": 46480
+    },
+    {
+      "epoch": 0.6119387271674475,
+      "grad_norm": 1.7487858533859253,
+      "learning_rate": 3.4558879259764506e-06,
+      "loss": 0.1679,
+      "step": 46500
+    },
+    {
+      "epoch": 0.6119387271674475,
+      "eval_loss": 0.20786046981811523,
+      "eval_runtime": 4.4633,
+      "eval_samples_per_second": 3.585,
+      "eval_steps_per_second": 1.792,
+      "step": 46500
+    },
+    {
+      "epoch": 0.6122019266199926,
+      "grad_norm": 1.9674745798110962,
+      "learning_rate": 3.4518346117686324e-06,
+      "loss": 0.1603,
+      "step": 46520
+    },
+    {
+      "epoch": 0.6124651260725378,
+      "grad_norm": 1.346795678138733,
+      "learning_rate": 3.4477824225486057e-06,
+      "loss": 0.1747,
+      "step": 46540
+    },
+    {
+      "epoch": 0.6127283255250829,
+      "grad_norm": 0.9670770764350891,
+      "learning_rate": 3.4437313612609314e-06,
+      "loss": 0.1839,
+      "step": 46560
+    },
+    {
+      "epoch": 0.6129915249776281,
+      "grad_norm": 1.4214129447937012,
+      "learning_rate": 3.439681430849348e-06,
+      "loss": 0.1791,
+      "step": 46580
+    },
+    {
+      "epoch": 0.6132547244301731,
+      "grad_norm": 1.4279203414916992,
+      "learning_rate": 3.4356326342567685e-06,
+      "loss": 0.1789,
+      "step": 46600
+    },
+    {
+      "epoch": 0.6135179238827183,
+      "grad_norm": 1.3152894973754883,
+      "learning_rate": 3.4315849744252925e-06,
+      "loss": 0.1673,
+      "step": 46620
+    },
+    {
+      "epoch": 0.6137811233352635,
+      "grad_norm": 2.178952217102051,
+      "learning_rate": 3.427538454296181e-06,
+      "loss": 0.1843,
+      "step": 46640
+    },
+    {
+      "epoch": 0.6140443227878086,
+      "grad_norm": 1.6584242582321167,
+      "learning_rate": 3.4234930768098772e-06,
+      "loss": 0.1847,
+      "step": 46660
+    },
+    {
+      "epoch": 0.6143075222403538,
+      "grad_norm": 1.9733648300170898,
+      "learning_rate": 3.419448844905988e-06,
+      "loss": 0.1682,
+      "step": 46680
+    },
+    {
+      "epoch": 0.6145707216928988,
+      "grad_norm": 1.5237318277359009,
+      "learning_rate": 3.4154057615232885e-06,
+      "loss": 0.1703,
+      "step": 46700
+    },
+    {
+      "epoch": 0.614833921145444,
+      "grad_norm": 1.1752742528915405,
+      "learning_rate": 3.411363829599722e-06,
+      "loss": 0.1803,
+      "step": 46720
+    },
+    {
+      "epoch": 0.6150971205979892,
+      "grad_norm": 1.7052134275436401,
+      "learning_rate": 3.4073230520723944e-06,
+      "loss": 0.1786,
+      "step": 46740
+    },
+    {
+      "epoch": 0.6153603200505343,
+      "grad_norm": 1.7607263326644897,
+      "learning_rate": 3.4032834318775666e-06,
+      "loss": 0.1755,
+      "step": 46760
+    },
+    {
+      "epoch": 0.6156235195030795,
+      "grad_norm": 2.3408114910125732,
+      "learning_rate": 3.3992449719506706e-06,
+      "loss": 0.1672,
+      "step": 46780
+    },
+    {
+      "epoch": 0.6158867189556245,
+      "grad_norm": 1.1930797100067139,
+      "learning_rate": 3.3952076752262807e-06,
+      "loss": 0.1774,
+      "step": 46800
+    },
+    {
+      "epoch": 0.6161499184081697,
+      "grad_norm": 1.5001200437545776,
+      "learning_rate": 3.3911715446381414e-06,
+      "loss": 0.1745,
+      "step": 46820
+    },
+    {
+      "epoch": 0.6164131178607148,
+      "grad_norm": 1.4914278984069824,
+      "learning_rate": 3.3871365831191373e-06,
+      "loss": 0.1761,
+      "step": 46840
+    },
+    {
+      "epoch": 0.61667631731326,
+      "grad_norm": 1.621340274810791,
+      "learning_rate": 3.383102793601308e-06,
+      "loss": 0.1776,
+      "step": 46860
+    },
+    {
+      "epoch": 0.6169395167658052,
+      "grad_norm": 1.3300524950027466,
+      "learning_rate": 3.3790701790158432e-06,
+      "loss": 0.1738,
+      "step": 46880
+    },
+    {
+      "epoch": 0.6172027162183502,
+      "grad_norm": 1.7594507932662964,
+      "learning_rate": 3.375038742293076e-06,
+      "loss": 0.1979,
+      "step": 46900
+    },
+    {
+      "epoch": 0.6174659156708954,
+      "grad_norm": 1.5927733182907104,
+      "learning_rate": 3.3710084863624857e-06,
+      "loss": 0.1772,
+      "step": 46920
+    },
+    {
+      "epoch": 0.6177291151234405,
+      "grad_norm": 1.9280422925949097,
+      "learning_rate": 3.3669794141526936e-06,
+      "loss": 0.1715,
+      "step": 46940
+    },
+    {
+      "epoch": 0.6179923145759857,
+      "grad_norm": 1.8575553894042969,
+      "learning_rate": 3.362951528591457e-06,
+      "loss": 0.1821,
+      "step": 46960
+    },
+    {
+      "epoch": 0.6182555140285309,
+      "grad_norm": 1.5526244640350342,
+      "learning_rate": 3.3589248326056773e-06,
+      "loss": 0.1533,
+      "step": 46980
+    },
+    {
+      "epoch": 0.618518713481076,
+      "grad_norm": 1.9379218816757202,
+      "learning_rate": 3.3548993291213848e-06,
+      "loss": 0.171,
+      "step": 47000
+    },
+    {
+      "epoch": 0.618518713481076,
+      "eval_loss": 0.2071094661951065,
+      "eval_runtime": 4.2726,
+      "eval_samples_per_second": 3.745,
+      "eval_steps_per_second": 1.872,
+      "step": 47000
+    },
+    {
+      "epoch": 0.6187819129336211,
+      "grad_norm": 1.473799467086792,
+      "learning_rate": 3.35087502106375e-06,
+      "loss": 0.1783,
+      "step": 47020
+    },
+    {
+      "epoch": 0.6190451123861662,
+      "grad_norm": 1.4250538349151611,
+      "learning_rate": 3.3468519113570687e-06,
+      "loss": 0.1837,
+      "step": 47040
+    },
+    {
+      "epoch": 0.6193083118387114,
+      "grad_norm": 1.7305587530136108,
+      "learning_rate": 3.3428300029247675e-06,
+      "loss": 0.1893,
+      "step": 47060
+    },
+    {
+      "epoch": 0.6195715112912565,
+      "grad_norm": 1.4048840999603271,
+      "learning_rate": 3.3388092986894037e-06,
+      "loss": 0.1891,
+      "step": 47080
+    },
+    {
+      "epoch": 0.6198347107438017,
+      "grad_norm": 1.3174940347671509,
+      "learning_rate": 3.334789801572656e-06,
+      "loss": 0.1761,
+      "step": 47100
+    },
+    {
+      "epoch": 0.6200979101963467,
+      "grad_norm": 1.5616940259933472,
+      "learning_rate": 3.330771514495325e-06,
+      "loss": 0.1744,
+      "step": 47120
+    },
+    {
+      "epoch": 0.6203611096488919,
+      "grad_norm": 1.6637237071990967,
+      "learning_rate": 3.3267544403773366e-06,
+      "loss": 0.1722,
+      "step": 47140
+    },
+    {
+      "epoch": 0.6206243091014371,
+      "grad_norm": 1.4821926355361938,
+      "learning_rate": 3.32273858213773e-06,
+      "loss": 0.1864,
+      "step": 47160
+    },
+    {
+      "epoch": 0.6208875085539822,
+      "grad_norm": 1.93845534324646,
+      "learning_rate": 3.3187239426946656e-06,
+      "loss": 0.1721,
+      "step": 47180
+    },
+    {
+      "epoch": 0.6211507080065274,
+      "grad_norm": 1.6154930591583252,
+      "learning_rate": 3.314710524965416e-06,
+      "loss": 0.1845,
+      "step": 47200
+    },
+    {
+      "epoch": 0.6214139074590724,
+      "grad_norm": 1.4771339893341064,
+      "learning_rate": 3.3106983318663617e-06,
+      "loss": 0.1662,
+      "step": 47220
+    },
+    {
+      "epoch": 0.6216771069116176,
+      "grad_norm": 1.486859917640686,
+      "learning_rate": 3.306687366313004e-06,
+      "loss": 0.1789,
+      "step": 47240
+    },
+    {
+      "epoch": 0.6219403063641628,
+      "grad_norm": 1.4396615028381348,
+      "learning_rate": 3.3026776312199393e-06,
+      "loss": 0.2018,
+      "step": 47260
+    },
+    {
+      "epoch": 0.6222035058167079,
+      "grad_norm": 2.100109815597534,
+      "learning_rate": 3.2986691295008823e-06,
+      "loss": 0.1834,
+      "step": 47280
+    },
+    {
+      "epoch": 0.6224667052692531,
+      "grad_norm": 1.9090827703475952,
+      "learning_rate": 3.2946618640686424e-06,
+      "loss": 0.1953,
+      "step": 47300
+    },
+    {
+      "epoch": 0.6227299047217982,
+      "grad_norm": 1.5369349718093872,
+      "learning_rate": 3.290655837835133e-06,
+      "loss": 0.1845,
+      "step": 47320
+    },
+    {
+      "epoch": 0.6229931041743433,
+      "grad_norm": 2.7796664237976074,
+      "learning_rate": 3.286651053711369e-06,
+      "loss": 0.1842,
+      "step": 47340
+    },
+    {
+      "epoch": 0.6232563036268884,
+      "grad_norm": 2.317786455154419,
+      "learning_rate": 3.282647514607464e-06,
+      "loss": 0.1833,
+      "step": 47360
+    },
+    {
+      "epoch": 0.6235195030794336,
+      "grad_norm": 1.6664737462997437,
+      "learning_rate": 3.278645223432617e-06,
+      "loss": 0.1723,
+      "step": 47380
+    },
+    {
+      "epoch": 0.6237827025319788,
+      "grad_norm": 1.5858327150344849,
+      "learning_rate": 3.274644183095137e-06,
+      "loss": 0.175,
+      "step": 47400
+    },
+    {
+      "epoch": 0.6240459019845239,
+      "grad_norm": 1.8453201055526733,
+      "learning_rate": 3.270644396502406e-06,
+      "loss": 0.1711,
+      "step": 47420
+    },
+    {
+      "epoch": 0.624309101437069,
+      "grad_norm": 1.194748878479004,
+      "learning_rate": 3.2666458665609102e-06,
+      "loss": 0.1655,
+      "step": 47440
+    },
+    {
+      "epoch": 0.6245723008896141,
+      "grad_norm": 1.3893145322799683,
+      "learning_rate": 3.2626485961762133e-06,
+      "loss": 0.1688,
+      "step": 47460
+    },
+    {
+      "epoch": 0.6248355003421593,
+      "grad_norm": 1.995284080505371,
+      "learning_rate": 3.2586525882529645e-06,
+      "loss": 0.1674,
+      "step": 47480
+    },
+    {
+      "epoch": 0.6250986997947044,
+      "grad_norm": 1.1936012506484985,
+      "learning_rate": 3.2546578456949014e-06,
+      "loss": 0.1825,
+      "step": 47500
+    },
+    {
+      "epoch": 0.6250986997947044,
+      "eval_loss": 0.20695281028747559,
+      "eval_runtime": 4.3442,
+      "eval_samples_per_second": 3.683,
+      "eval_steps_per_second": 1.842,
+      "step": 47500
+    },
+    {
+      "epoch": 0.6253618992472496,
+      "grad_norm": 1.5718638896942139,
+      "learning_rate": 3.250664371404835e-06,
+      "loss": 0.1699,
+      "step": 47520
+    },
+    {
+      "epoch": 0.6256250986997947,
+      "grad_norm": 1.9470757246017456,
+      "learning_rate": 3.2466721682846603e-06,
+      "loss": 0.1799,
+      "step": 47540
+    },
+    {
+      "epoch": 0.6258882981523398,
+      "grad_norm": 1.5858714580535889,
+      "learning_rate": 3.2426812392353457e-06,
+      "loss": 0.1711,
+      "step": 47560
+    },
+    {
+      "epoch": 0.626151497604885,
+      "grad_norm": 1.8288679122924805,
+      "learning_rate": 3.2386915871569337e-06,
+      "loss": 0.1701,
+      "step": 47580
+    },
+    {
+      "epoch": 0.6264146970574301,
+      "grad_norm": 1.4530377388000488,
+      "learning_rate": 3.2347032149485402e-06,
+      "loss": 0.1687,
+      "step": 47600
+    },
+    {
+      "epoch": 0.6266778965099753,
+      "grad_norm": 1.831730842590332,
+      "learning_rate": 3.2307161255083513e-06,
+      "loss": 0.1846,
+      "step": 47620
+    },
+    {
+      "epoch": 0.6269410959625203,
+      "grad_norm": 2.2925286293029785,
+      "learning_rate": 3.226730321733617e-06,
+      "loss": 0.1685,
+      "step": 47640
+    },
+    {
+      "epoch": 0.6272042954150655,
+      "grad_norm": 1.7345302104949951,
+      "learning_rate": 3.2227458065206597e-06,
+      "loss": 0.1871,
+      "step": 47660
+    },
+    {
+      "epoch": 0.6274674948676107,
+      "grad_norm": 1.6631433963775635,
+      "learning_rate": 3.2187625827648585e-06,
+      "loss": 0.1607,
+      "step": 47680
+    },
+    {
+      "epoch": 0.6277306943201558,
+      "grad_norm": 1.8012304306030273,
+      "learning_rate": 3.214780653360661e-06,
+      "loss": 0.1699,
+      "step": 47700
+    },
+    {
+      "epoch": 0.627993893772701,
+      "grad_norm": 1.9489649534225464,
+      "learning_rate": 3.210800021201568e-06,
+      "loss": 0.1889,
+      "step": 47720
+    },
+    {
+      "epoch": 0.6282570932252461,
+      "grad_norm": 1.3559623956680298,
+      "learning_rate": 3.2068206891801414e-06,
+      "loss": 0.186,
+      "step": 47740
+    },
+    {
+      "epoch": 0.6285202926777912,
+      "grad_norm": 1.8115894794464111,
+      "learning_rate": 3.202842660187997e-06,
+      "loss": 0.1673,
+      "step": 47760
+    },
+    {
+      "epoch": 0.6287834921303364,
+      "grad_norm": 1.2197420597076416,
+      "learning_rate": 3.198865937115804e-06,
+      "loss": 0.1815,
+      "step": 47780
+    },
+    {
+      "epoch": 0.6290466915828815,
+      "grad_norm": 1.5773674249649048,
+      "learning_rate": 3.1948905228532843e-06,
+      "loss": 0.1725,
+      "step": 47800
+    },
+    {
+      "epoch": 0.6293098910354267,
+      "grad_norm": 2.314321517944336,
+      "learning_rate": 3.1909164202892063e-06,
+      "loss": 0.163,
+      "step": 47820
+    },
+    {
+      "epoch": 0.6295730904879718,
+      "grad_norm": 1.5918972492218018,
+      "learning_rate": 3.1869436323113822e-06,
+      "loss": 0.1959,
+      "step": 47840
+    },
+    {
+      "epoch": 0.6298362899405169,
+      "grad_norm": 1.6597886085510254,
+      "learning_rate": 3.18297216180668e-06,
+      "loss": 0.166,
+      "step": 47860
+    },
+    {
+      "epoch": 0.630099489393062,
+      "grad_norm": 1.829323172569275,
+      "learning_rate": 3.1790020116609976e-06,
+      "loss": 0.1648,
+      "step": 47880
+    },
+    {
+      "epoch": 0.6303626888456072,
+      "grad_norm": 1.4335193634033203,
+      "learning_rate": 3.1750331847592797e-06,
+      "loss": 0.1838,
+      "step": 47900
+    },
+    {
+      "epoch": 0.6306258882981524,
+      "grad_norm": 1.848875641822815,
+      "learning_rate": 3.1710656839855103e-06,
+      "loss": 0.1781,
+      "step": 47920
+    },
+    {
+      "epoch": 0.6308890877506975,
+      "grad_norm": 1.4996896982192993,
+      "learning_rate": 3.167099512222708e-06,
+      "loss": 0.1864,
+      "step": 47940
+    },
+    {
+      "epoch": 0.6311522872032426,
+      "grad_norm": 1.4469743967056274,
+      "learning_rate": 3.163134672352925e-06,
+      "loss": 0.179,
+      "step": 47960
+    },
+    {
+      "epoch": 0.6314154866557877,
+      "grad_norm": 1.6737077236175537,
+      "learning_rate": 3.159171167257248e-06,
+      "loss": 0.1737,
+      "step": 47980
+    },
+    {
+      "epoch": 0.6316786861083329,
+      "grad_norm": 1.2042763233184814,
+      "learning_rate": 3.1552089998157897e-06,
+      "loss": 0.1831,
+      "step": 48000
+    },
+    {
+      "epoch": 0.6316786861083329,
+      "eval_loss": 0.21535879373550415,
+      "eval_runtime": 4.4314,
+      "eval_samples_per_second": 3.611,
+      "eval_steps_per_second": 1.805,
+      "step": 48000
+    },
+    {
+      "epoch": 0.631941885560878,
+      "grad_norm": 1.8067047595977783,
+      "learning_rate": 3.1512481729076973e-06,
+      "loss": 0.1781,
+      "step": 48020
+    },
+    {
+      "epoch": 0.6322050850134232,
+      "grad_norm": 1.7278237342834473,
+      "learning_rate": 3.1472886894111344e-06,
+      "loss": 0.1901,
+      "step": 48040
+    },
+    {
+      "epoch": 0.6324682844659683,
+      "grad_norm": 1.7038389444351196,
+      "learning_rate": 3.143330552203301e-06,
+      "loss": 0.188,
+      "step": 48060
+    },
+    {
+      "epoch": 0.6327314839185134,
+      "grad_norm": 1.3073911666870117,
+      "learning_rate": 3.1393737641604074e-06,
+      "loss": 0.1778,
+      "step": 48080
+    },
+    {
+      "epoch": 0.6329946833710586,
+      "grad_norm": 1.998342752456665,
+      "learning_rate": 3.1354183281576867e-06,
+      "loss": 0.1711,
+      "step": 48100
+    },
+    {
+      "epoch": 0.6332578828236037,
+      "grad_norm": 1.3511220216751099,
+      "learning_rate": 3.131464247069394e-06,
+      "loss": 0.1862,
+      "step": 48120
+    },
+    {
+      "epoch": 0.6335210822761489,
+      "grad_norm": 2.4194061756134033,
+      "learning_rate": 3.127511523768795e-06,
+      "loss": 0.1682,
+      "step": 48140
+    },
+    {
+      "epoch": 0.633784281728694,
+      "grad_norm": 2.0113093852996826,
+      "learning_rate": 3.1235601611281703e-06,
+      "loss": 0.1751,
+      "step": 48160
+    },
+    {
+      "epoch": 0.6340474811812391,
+      "grad_norm": 2.6837852001190186,
+      "learning_rate": 3.1196101620188126e-06,
+      "loss": 0.1795,
+      "step": 48180
+    },
+    {
+      "epoch": 0.6343106806337843,
+      "grad_norm": 1.7768820524215698,
+      "learning_rate": 3.1156615293110214e-06,
+      "loss": 0.1761,
+      "step": 48200
+    },
+    {
+      "epoch": 0.6345738800863294,
+      "grad_norm": 1.4697308540344238,
+      "learning_rate": 3.1117142658741073e-06,
+      "loss": 0.1784,
+      "step": 48220
+    },
+    {
+      "epoch": 0.6348370795388746,
+      "grad_norm": 1.5880018472671509,
+      "learning_rate": 3.107768374576383e-06,
+      "loss": 0.1701,
+      "step": 48240
+    },
+    {
+      "epoch": 0.6351002789914197,
+      "grad_norm": 2.045236587524414,
+      "learning_rate": 3.103823858285161e-06,
+      "loss": 0.151,
+      "step": 48260
+    },
+    {
+      "epoch": 0.6353634784439648,
+      "grad_norm": 1.582715392112732,
+      "learning_rate": 3.0998807198667635e-06,
+      "loss": 0.1617,
+      "step": 48280
+    },
+    {
+      "epoch": 0.63562667789651,
+      "grad_norm": 1.421105980873108,
+      "learning_rate": 3.0959389621865022e-06,
+      "loss": 0.1692,
+      "step": 48300
+    },
+    {
+      "epoch": 0.6358898773490551,
+      "grad_norm": 1.7165493965148926,
+      "learning_rate": 3.091998588108691e-06,
+      "loss": 0.1654,
+      "step": 48320
+    },
+    {
+      "epoch": 0.6361530768016003,
+      "grad_norm": 1.2492643594741821,
+      "learning_rate": 3.088059600496637e-06,
+      "loss": 0.1757,
+      "step": 48340
+    },
+    {
+      "epoch": 0.6364162762541454,
+      "grad_norm": 1.7100573778152466,
+      "learning_rate": 3.0841220022126394e-06,
+      "loss": 0.1876,
+      "step": 48360
+    },
+    {
+      "epoch": 0.6366794757066905,
+      "grad_norm": 1.5020900964736938,
+      "learning_rate": 3.080185796117988e-06,
+      "loss": 0.1586,
+      "step": 48380
+    },
+    {
+      "epoch": 0.6369426751592356,
+      "grad_norm": 1.4372478723526,
+      "learning_rate": 3.076250985072961e-06,
+      "loss": 0.1684,
+      "step": 48400
+    },
+    {
+      "epoch": 0.6372058746117808,
+      "grad_norm": 1.3610224723815918,
+      "learning_rate": 3.0723175719368236e-06,
+      "loss": 0.1926,
+      "step": 48420
+    },
+    {
+      "epoch": 0.637469074064326,
+      "grad_norm": 1.7022435665130615,
+      "learning_rate": 3.068385559567826e-06,
+      "loss": 0.178,
+      "step": 48440
+    },
+    {
+      "epoch": 0.6377322735168711,
+      "grad_norm": 1.0407930612564087,
+      "learning_rate": 3.0644549508231935e-06,
+      "loss": 0.182,
+      "step": 48460
+    },
+    {
+      "epoch": 0.6379954729694163,
+      "grad_norm": 1.1613706350326538,
+      "learning_rate": 3.060525748559144e-06,
+      "loss": 0.1767,
+      "step": 48480
+    },
+    {
+      "epoch": 0.6382586724219613,
+      "grad_norm": 1.2593967914581299,
+      "learning_rate": 3.0565979556308612e-06,
+      "loss": 0.1759,
+      "step": 48500
+    },
+    {
+      "epoch": 0.6382586724219613,
+      "eval_loss": 0.208587646484375,
+      "eval_runtime": 4.3405,
+      "eval_samples_per_second": 3.686,
+      "eval_steps_per_second": 1.843,
+      "step": 48500
+    },
+    {
+      "epoch": 0.6385218718745065,
+      "grad_norm": 1.2589281797409058,
+      "learning_rate": 3.052671574892512e-06,
+      "loss": 0.1774,
+      "step": 48520
+    },
+    {
+      "epoch": 0.6387850713270516,
+      "grad_norm": 1.3630975484848022,
+      "learning_rate": 3.0487466091972347e-06,
+      "loss": 0.1651,
+      "step": 48540
+    },
+    {
+      "epoch": 0.6390482707795968,
+      "grad_norm": 1.487373948097229,
+      "learning_rate": 3.0448230613971388e-06,
+      "loss": 0.1753,
+      "step": 48560
+    },
+    {
+      "epoch": 0.639311470232142,
+      "grad_norm": 1.6876944303512573,
+      "learning_rate": 3.040900934343306e-06,
+      "loss": 0.1861,
+      "step": 48580
+    },
+    {
+      "epoch": 0.639574669684687,
+      "grad_norm": 1.5678824186325073,
+      "learning_rate": 3.0369802308857834e-06,
+      "loss": 0.1796,
+      "step": 48600
+    },
+    {
+      "epoch": 0.6398378691372322,
+      "grad_norm": 1.51077401638031,
+      "learning_rate": 3.0330609538735823e-06,
+      "loss": 0.1722,
+      "step": 48620
+    },
+    {
+      "epoch": 0.6401010685897773,
+      "grad_norm": 1.836600661277771,
+      "learning_rate": 3.0291431061546816e-06,
+      "loss": 0.1806,
+      "step": 48640
+    },
+    {
+      "epoch": 0.6403642680423225,
+      "grad_norm": 1.2452199459075928,
+      "learning_rate": 3.025226690576018e-06,
+      "loss": 0.1826,
+      "step": 48660
+    },
+    {
+      "epoch": 0.6406274674948677,
+      "grad_norm": 1.7011995315551758,
+      "learning_rate": 3.021311709983491e-06,
+      "loss": 0.1759,
+      "step": 48680
+    },
+    {
+      "epoch": 0.6408906669474127,
+      "grad_norm": 1.1921391487121582,
+      "learning_rate": 3.017398167221952e-06,
+      "loss": 0.1698,
+      "step": 48700
+    },
+    {
+      "epoch": 0.6411538663999579,
+      "grad_norm": 1.294022560119629,
+      "learning_rate": 3.0134860651352116e-06,
+      "loss": 0.1905,
+      "step": 48720
+    },
+    {
+      "epoch": 0.641417065852503,
+      "grad_norm": 1.9401922225952148,
+      "learning_rate": 3.0095754065660343e-06,
+      "loss": 0.1749,
+      "step": 48740
+    },
+    {
+      "epoch": 0.6416802653050482,
+      "grad_norm": 1.322553277015686,
+      "learning_rate": 3.0056661943561327e-06,
+      "loss": 0.1759,
+      "step": 48760
+    },
+    {
+      "epoch": 0.6419434647575933,
+      "grad_norm": 1.6119738817214966,
+      "learning_rate": 3.0017584313461695e-06,
+      "loss": 0.1757,
+      "step": 48780
+    },
+    {
+      "epoch": 0.6422066642101384,
+      "grad_norm": 1.2031006813049316,
+      "learning_rate": 2.997852120375756e-06,
+      "loss": 0.1591,
+      "step": 48800
+    },
+    {
+      "epoch": 0.6424698636626835,
+      "grad_norm": 1.5688413381576538,
+      "learning_rate": 2.9939472642834445e-06,
+      "loss": 0.1658,
+      "step": 48820
+    },
+    {
+      "epoch": 0.6427330631152287,
+      "grad_norm": 1.6775826215744019,
+      "learning_rate": 2.9900438659067347e-06,
+      "loss": 0.167,
+      "step": 48840
+    },
+    {
+      "epoch": 0.6429962625677739,
+      "grad_norm": 2.286435127258301,
+      "learning_rate": 2.9861419280820658e-06,
+      "loss": 0.1588,
+      "step": 48860
+    },
+    {
+      "epoch": 0.643259462020319,
+      "grad_norm": 1.9200060367584229,
+      "learning_rate": 2.9822414536448097e-06,
+      "loss": 0.1842,
+      "step": 48880
+    },
+    {
+      "epoch": 0.6435226614728642,
+      "grad_norm": 1.3142998218536377,
+      "learning_rate": 2.9783424454292865e-06,
+      "loss": 0.1552,
+      "step": 48900
+    },
+    {
+      "epoch": 0.6437858609254092,
+      "grad_norm": 1.8372858762741089,
+      "learning_rate": 2.974444906268738e-06,
+      "loss": 0.1708,
+      "step": 48920
+    },
+    {
+      "epoch": 0.6440490603779544,
+      "grad_norm": 1.7279396057128906,
+      "learning_rate": 2.9705488389953506e-06,
+      "loss": 0.1798,
+      "step": 48940
+    },
+    {
+      "epoch": 0.6443122598304996,
+      "grad_norm": 1.1070308685302734,
+      "learning_rate": 2.966654246440232e-06,
+      "loss": 0.1803,
+      "step": 48960
+    },
+    {
+      "epoch": 0.6445754592830447,
+      "grad_norm": 1.867307424545288,
+      "learning_rate": 2.9627611314334216e-06,
+      "loss": 0.1556,
+      "step": 48980
+    },
+    {
+      "epoch": 0.6448386587355899,
+      "grad_norm": 1.3768038749694824,
+      "learning_rate": 2.9588694968038856e-06,
+      "loss": 0.1713,
+      "step": 49000
+    },
+    {
+      "epoch": 0.6448386587355899,
+      "eval_loss": 0.20705682039260864,
+      "eval_runtime": 4.201,
+      "eval_samples_per_second": 3.809,
+      "eval_steps_per_second": 1.904,
+      "step": 49000
+    },
+    {
+      "epoch": 0.6451018581881349,
+      "grad_norm": 1.9192309379577637,
+      "learning_rate": 2.954979345379516e-06,
+      "loss": 0.1657,
+      "step": 49020
+    },
+    {
+      "epoch": 0.6453650576406801,
+      "grad_norm": 2.6102583408355713,
+      "learning_rate": 2.95109067998712e-06,
+      "loss": 0.1739,
+      "step": 49040
+    },
+    {
+      "epoch": 0.6456282570932252,
+      "grad_norm": 2.059788227081299,
+      "learning_rate": 2.947203503452437e-06,
+      "loss": 0.19,
+      "step": 49060
+    },
+    {
+      "epoch": 0.6458914565457704,
+      "grad_norm": 1.7807469367980957,
+      "learning_rate": 2.94331781860011e-06,
+      "loss": 0.1731,
+      "step": 49080
+    },
+    {
+      "epoch": 0.6461546559983156,
+      "grad_norm": 1.2437772750854492,
+      "learning_rate": 2.939433628253713e-06,
+      "loss": 0.1802,
+      "step": 49100
+    },
+    {
+      "epoch": 0.6464178554508606,
+      "grad_norm": 1.7560604810714722,
+      "learning_rate": 2.9355509352357207e-06,
+      "loss": 0.1896,
+      "step": 49120
+    },
+    {
+      "epoch": 0.6466810549034058,
+      "grad_norm": 1.8712421655654907,
+      "learning_rate": 2.931669742367528e-06,
+      "loss": 0.1738,
+      "step": 49140
+    },
+    {
+      "epoch": 0.6469442543559509,
+      "grad_norm": 1.6922128200531006,
+      "learning_rate": 2.9277900524694376e-06,
+      "loss": 0.1691,
+      "step": 49160
+    },
+    {
+      "epoch": 0.6472074538084961,
+      "grad_norm": 1.8487828969955444,
+      "learning_rate": 2.923911868360658e-06,
+      "loss": 0.1751,
+      "step": 49180
+    },
+    {
+      "epoch": 0.6474706532610413,
+      "grad_norm": 2.069378137588501,
+      "learning_rate": 2.9200351928593083e-06,
+      "loss": 0.1706,
+      "step": 49200
+    },
+    {
+      "epoch": 0.6477338527135864,
+      "grad_norm": 1.972724437713623,
+      "learning_rate": 2.9161600287824073e-06,
+      "loss": 0.1843,
+      "step": 49220
+    },
+    {
+      "epoch": 0.6479970521661315,
+      "grad_norm": 1.5123308897018433,
+      "learning_rate": 2.9122863789458728e-06,
+      "loss": 0.1775,
+      "step": 49240
+    },
+    {
+      "epoch": 0.6482602516186766,
+      "grad_norm": 1.6153842210769653,
+      "learning_rate": 2.9084142461645317e-06,
+      "loss": 0.1956,
+      "step": 49260
+    },
+    {
+      "epoch": 0.6485234510712218,
+      "grad_norm": 1.5204365253448486,
+      "learning_rate": 2.904543633252099e-06,
+      "loss": 0.1901,
+      "step": 49280
+    },
+    {
+      "epoch": 0.6487866505237669,
+      "grad_norm": 1.7582322359085083,
+      "learning_rate": 2.900674543021191e-06,
+      "loss": 0.1848,
+      "step": 49300
+    },
+    {
+      "epoch": 0.6490498499763121,
+      "grad_norm": 1.2047014236450195,
+      "learning_rate": 2.8968069782833152e-06,
+      "loss": 0.1783,
+      "step": 49320
+    },
+    {
+      "epoch": 0.6493130494288571,
+      "grad_norm": 2.4559454917907715,
+      "learning_rate": 2.892940941848869e-06,
+      "loss": 0.1777,
+      "step": 49340
+    },
+    {
+      "epoch": 0.6495762488814023,
+      "grad_norm": 1.5397212505340576,
+      "learning_rate": 2.889076436527145e-06,
+      "loss": 0.1598,
+      "step": 49360
+    },
+    {
+      "epoch": 0.6498394483339475,
+      "grad_norm": 1.16042959690094,
+      "learning_rate": 2.8852134651263192e-06,
+      "loss": 0.1734,
+      "step": 49380
+    },
+    {
+      "epoch": 0.6501026477864926,
+      "grad_norm": 1.5392036437988281,
+      "learning_rate": 2.881352030453448e-06,
+      "loss": 0.1772,
+      "step": 49400
+    },
+    {
+      "epoch": 0.6503658472390378,
+      "grad_norm": 1.9292936325073242,
+      "learning_rate": 2.8774921353144812e-06,
+      "loss": 0.1789,
+      "step": 49420
+    },
+    {
+      "epoch": 0.6506290466915828,
+      "grad_norm": 1.9955825805664062,
+      "learning_rate": 2.8736337825142445e-06,
+      "loss": 0.1826,
+      "step": 49440
+    },
+    {
+      "epoch": 0.650892246144128,
+      "grad_norm": 1.1549440622329712,
+      "learning_rate": 2.8697769748564432e-06,
+      "loss": 0.1645,
+      "step": 49460
+    },
+    {
+      "epoch": 0.6511554455966732,
+      "grad_norm": 1.2502585649490356,
+      "learning_rate": 2.865921715143658e-06,
+      "loss": 0.2014,
+      "step": 49480
+    },
+    {
+      "epoch": 0.6514186450492183,
+      "grad_norm": 1.716629147529602,
+      "learning_rate": 2.8620680061773474e-06,
+      "loss": 0.1832,
+      "step": 49500
+    },
+    {
+      "epoch": 0.6514186450492183,
+      "eval_loss": 0.20339149236679077,
+      "eval_runtime": 4.1979,
+      "eval_samples_per_second": 3.811,
+      "eval_steps_per_second": 1.906,
+      "step": 49500
+    },
+    {
+      "epoch": 0.6516818445017635,
+      "grad_norm": 1.4478518962860107,
+      "learning_rate": 2.858215850757845e-06,
+      "loss": 0.184,
+      "step": 49520
+    },
+    {
+      "epoch": 0.6519450439543085,
+      "grad_norm": 1.8047629594802856,
+      "learning_rate": 2.8543652516843535e-06,
+      "loss": 0.1636,
+      "step": 49540
+    },
+    {
+      "epoch": 0.6522082434068537,
+      "grad_norm": 1.7045538425445557,
+      "learning_rate": 2.850516211754937e-06,
+      "loss": 0.1768,
+      "step": 49560
+    },
+    {
+      "epoch": 0.6524714428593988,
+      "grad_norm": 1.5404870510101318,
+      "learning_rate": 2.8466687337665413e-06,
+      "loss": 0.174,
+      "step": 49580
+    },
+    {
+      "epoch": 0.652734642311944,
+      "grad_norm": 1.2940205335617065,
+      "learning_rate": 2.8428228205149646e-06,
+      "loss": 0.181,
+      "step": 49600
+    },
+    {
+      "epoch": 0.6529978417644892,
+      "grad_norm": 1.5318336486816406,
+      "learning_rate": 2.8389784747948785e-06,
+      "loss": 0.1654,
+      "step": 49620
+    },
+    {
+      "epoch": 0.6532610412170343,
+      "grad_norm": 1.3763635158538818,
+      "learning_rate": 2.8351356993998057e-06,
+      "loss": 0.1797,
+      "step": 49640
+    },
+    {
+      "epoch": 0.6535242406695794,
+      "grad_norm": 1.3653596639633179,
+      "learning_rate": 2.8312944971221313e-06,
+      "loss": 0.1754,
+      "step": 49660
+    },
+    {
+      "epoch": 0.6537874401221245,
+      "grad_norm": 1.7202982902526855,
+      "learning_rate": 2.8274548707531017e-06,
+      "loss": 0.1989,
+      "step": 49680
+    },
+    {
+      "epoch": 0.6540506395746697,
+      "grad_norm": 1.2771930694580078,
+      "learning_rate": 2.823616823082814e-06,
+      "loss": 0.1983,
+      "step": 49700
+    },
+    {
+      "epoch": 0.6543138390272148,
+      "grad_norm": 1.6970371007919312,
+      "learning_rate": 2.819780356900218e-06,
+      "loss": 0.1809,
+      "step": 49720
+    },
+    {
+      "epoch": 0.65457703847976,
+      "grad_norm": 1.3386476039886475,
+      "learning_rate": 2.8159454749931152e-06,
+      "loss": 0.1821,
+      "step": 49740
+    },
+    {
+      "epoch": 0.6548402379323051,
+      "grad_norm": 1.6036735773086548,
+      "learning_rate": 2.812112180148154e-06,
+      "loss": 0.171,
+      "step": 49760
+    },
+    {
+      "epoch": 0.6551034373848502,
+      "grad_norm": 2.1867430210113525,
+      "learning_rate": 2.808280475150835e-06,
+      "loss": 0.173,
+      "step": 49780
+    },
+    {
+      "epoch": 0.6553666368373954,
+      "grad_norm": 1.6054679155349731,
+      "learning_rate": 2.8044503627854974e-06,
+      "loss": 0.1712,
+      "step": 49800
+    },
+    {
+      "epoch": 0.6556298362899405,
+      "grad_norm": 1.1797575950622559,
+      "learning_rate": 2.8006218458353264e-06,
+      "loss": 0.172,
+      "step": 49820
+    },
+    {
+      "epoch": 0.6558930357424857,
+      "grad_norm": 1.427444577217102,
+      "learning_rate": 2.796794927082347e-06,
+      "loss": 0.1886,
+      "step": 49840
+    },
+    {
+      "epoch": 0.6561562351950307,
+      "grad_norm": 1.4769588708877563,
+      "learning_rate": 2.792969609307422e-06,
+      "loss": 0.1699,
+      "step": 49860
+    },
+    {
+      "epoch": 0.6564194346475759,
+      "grad_norm": 1.5464262962341309,
+      "learning_rate": 2.7891458952902526e-06,
+      "loss": 0.1751,
+      "step": 49880
+    },
+    {
+      "epoch": 0.6566826341001211,
+      "grad_norm": 1.495956540107727,
+      "learning_rate": 2.7853237878093733e-06,
+      "loss": 0.1671,
+      "step": 49900
+    },
+    {
+      "epoch": 0.6569458335526662,
+      "grad_norm": 1.55937922000885,
+      "learning_rate": 2.781503289642149e-06,
+      "loss": 0.1791,
+      "step": 49920
+    },
+    {
+      "epoch": 0.6572090330052114,
+      "grad_norm": 1.7281901836395264,
+      "learning_rate": 2.7776844035647833e-06,
+      "loss": 0.1716,
+      "step": 49940
+    },
+    {
+      "epoch": 0.6574722324577564,
+      "grad_norm": 1.5152987241744995,
+      "learning_rate": 2.7738671323522997e-06,
+      "loss": 0.1638,
+      "step": 49960
+    },
+    {
+      "epoch": 0.6577354319103016,
+      "grad_norm": 1.1514426469802856,
+      "learning_rate": 2.7700514787785515e-06,
+      "loss": 0.1885,
+      "step": 49980
+    },
+    {
+      "epoch": 0.6579986313628468,
+      "grad_norm": 2.019207715988159,
+      "learning_rate": 2.7662374456162174e-06,
+      "loss": 0.1851,
+      "step": 50000
+    },
+    {
+      "epoch": 0.6579986313628468,
+      "eval_loss": 0.20929107069969177,
+      "eval_runtime": 4.3355,
+      "eval_samples_per_second": 3.69,
+      "eval_steps_per_second": 1.845,
+      "step": 50000
+    },
+    {
+      "epoch": 0.6582618308153919,
+      "grad_norm": 2.040736436843872,
+      "learning_rate": 2.762425035636794e-06,
+      "loss": 0.167,
+      "step": 50020
+    },
+    {
+      "epoch": 0.6585250302679371,
+      "grad_norm": 1.6770601272583008,
+      "learning_rate": 2.7586142516106094e-06,
+      "loss": 0.1743,
+      "step": 50040
+    },
+    {
+      "epoch": 0.6587882297204822,
+      "grad_norm": 1.3764272928237915,
+      "learning_rate": 2.7548050963067946e-06,
+      "loss": 0.1648,
+      "step": 50060
+    },
+    {
+      "epoch": 0.6590514291730273,
+      "grad_norm": 1.83476984500885,
+      "learning_rate": 2.7509975724933104e-06,
+      "loss": 0.1608,
+      "step": 50080
+    },
+    {
+      "epoch": 0.6593146286255724,
+      "grad_norm": 1.1793839931488037,
+      "learning_rate": 2.7471916829369263e-06,
+      "loss": 0.1822,
+      "step": 50100
+    },
+    {
+      "epoch": 0.6595778280781176,
+      "grad_norm": 1.2250794172286987,
+      "learning_rate": 2.743387430403224e-06,
+      "loss": 0.1711,
+      "step": 50120
+    },
+    {
+      "epoch": 0.6598410275306628,
+      "grad_norm": 1.7892848253250122,
+      "learning_rate": 2.7395848176565965e-06,
+      "loss": 0.1826,
+      "step": 50140
+    },
+    {
+      "epoch": 0.6601042269832079,
+      "grad_norm": 1.4487804174423218,
+      "learning_rate": 2.735783847460246e-06,
+      "loss": 0.1869,
+      "step": 50160
+    },
+    {
+      "epoch": 0.660367426435753,
+      "grad_norm": 1.4488286972045898,
+      "learning_rate": 2.7319845225761775e-06,
+      "loss": 0.1581,
+      "step": 50180
+    },
+    {
+      "epoch": 0.6606306258882981,
+      "grad_norm": 1.6842427253723145,
+      "learning_rate": 2.7281868457652104e-06,
+      "loss": 0.1563,
+      "step": 50200
+    },
+    {
+      "epoch": 0.6608938253408433,
+      "grad_norm": 1.6623625755310059,
+      "learning_rate": 2.724390819786951e-06,
+      "loss": 0.165,
+      "step": 50220
+    },
+    {
+      "epoch": 0.6611570247933884,
+      "grad_norm": 1.295477271080017,
+      "learning_rate": 2.7205964473998205e-06,
+      "loss": 0.1804,
+      "step": 50240
+    },
+    {
+      "epoch": 0.6614202242459336,
+      "grad_norm": 1.6778920888900757,
+      "learning_rate": 2.71680373136103e-06,
+      "loss": 0.1826,
+      "step": 50260
+    },
+    {
+      "epoch": 0.6616834236984787,
+      "grad_norm": 1.7189381122589111,
+      "learning_rate": 2.71301267442659e-06,
+      "loss": 0.1688,
+      "step": 50280
+    },
+    {
+      "epoch": 0.6619466231510238,
+      "grad_norm": 1.5713120698928833,
+      "learning_rate": 2.7092232793513053e-06,
+      "loss": 0.1852,
+      "step": 50300
+    },
+    {
+      "epoch": 0.662209822603569,
+      "grad_norm": 1.73651123046875,
+      "learning_rate": 2.705435548888769e-06,
+      "loss": 0.1785,
+      "step": 50320
+    },
+    {
+      "epoch": 0.6624730220561141,
+      "grad_norm": 1.6175687313079834,
+      "learning_rate": 2.701649485791374e-06,
+      "loss": 0.1673,
+      "step": 50340
+    },
+    {
+      "epoch": 0.6627362215086593,
+      "grad_norm": 0.8876596689224243,
+      "learning_rate": 2.697865092810293e-06,
+      "loss": 0.1737,
+      "step": 50360
+    },
+    {
+      "epoch": 0.6629994209612045,
+      "grad_norm": 2.0122549533843994,
+      "learning_rate": 2.694082372695488e-06,
+      "loss": 0.1759,
+      "step": 50380
+    },
+    {
+      "epoch": 0.6632626204137495,
+      "grad_norm": 1.522867202758789,
+      "learning_rate": 2.6903013281957057e-06,
+      "loss": 0.1749,
+      "step": 50400
+    },
+    {
+      "epoch": 0.6635258198662947,
+      "grad_norm": 1.9850056171417236,
+      "learning_rate": 2.6865219620584744e-06,
+      "loss": 0.1607,
+      "step": 50420
+    },
+    {
+      "epoch": 0.6637890193188398,
+      "grad_norm": 1.9484671354293823,
+      "learning_rate": 2.6827442770301037e-06,
+      "loss": 0.1996,
+      "step": 50440
+    },
+    {
+      "epoch": 0.664052218771385,
+      "grad_norm": 1.7641352415084839,
+      "learning_rate": 2.6789682758556806e-06,
+      "loss": 0.1751,
+      "step": 50460
+    },
+    {
+      "epoch": 0.6643154182239301,
+      "grad_norm": 1.0869990587234497,
+      "learning_rate": 2.675193961279068e-06,
+      "loss": 0.1629,
+      "step": 50480
+    },
+    {
+      "epoch": 0.6645786176764752,
+      "grad_norm": 1.9408844709396362,
+      "learning_rate": 2.6714213360429076e-06,
+      "loss": 0.1695,
+      "step": 50500
+    },
+    {
+      "epoch": 0.6645786176764752,
+      "eval_loss": 0.20975002646446228,
+      "eval_runtime": 4.2201,
+      "eval_samples_per_second": 3.791,
+      "eval_steps_per_second": 1.896,
+      "step": 50500
+    },
+    {
+      "epoch": 0.6648418171290204,
+      "grad_norm": 1.6416517496109009,
+      "learning_rate": 2.6676504028886076e-06,
+      "loss": 0.1775,
+      "step": 50520
+    },
+    {
+      "epoch": 0.6651050165815655,
+      "grad_norm": 1.5344107151031494,
+      "learning_rate": 2.6638811645563506e-06,
+      "loss": 0.1746,
+      "step": 50540
+    },
+    {
+      "epoch": 0.6653682160341107,
+      "grad_norm": 1.533766746520996,
+      "learning_rate": 2.6601136237850853e-06,
+      "loss": 0.1748,
+      "step": 50560
+    },
+    {
+      "epoch": 0.6656314154866558,
+      "grad_norm": 1.4750057458877563,
+      "learning_rate": 2.656347783312526e-06,
+      "loss": 0.2006,
+      "step": 50580
+    },
+    {
+      "epoch": 0.6658946149392009,
+      "grad_norm": 1.4692058563232422,
+      "learning_rate": 2.6525836458751588e-06,
+      "loss": 0.2021,
+      "step": 50600
+    },
+    {
+      "epoch": 0.666157814391746,
+      "grad_norm": 1.334357738494873,
+      "learning_rate": 2.6488212142082236e-06,
+      "loss": 0.1702,
+      "step": 50620
+    },
+    {
+      "epoch": 0.6664210138442912,
+      "grad_norm": 1.2043869495391846,
+      "learning_rate": 2.645060491045721e-06,
+      "loss": 0.1783,
+      "step": 50640
+    },
+    {
+      "epoch": 0.6666842132968364,
+      "grad_norm": 1.504189372062683,
+      "learning_rate": 2.6413014791204176e-06,
+      "loss": 0.1828,
+      "step": 50660
+    },
+    {
+      "epoch": 0.6669474127493815,
+      "grad_norm": 1.5764122009277344,
+      "learning_rate": 2.6375441811638304e-06,
+      "loss": 0.1733,
+      "step": 50680
+    },
+    {
+      "epoch": 0.6672106122019266,
+      "grad_norm": 1.5389775037765503,
+      "learning_rate": 2.6337885999062324e-06,
+      "loss": 0.1709,
+      "step": 50700
+    },
+    {
+      "epoch": 0.6674738116544717,
+      "grad_norm": 1.2762243747711182,
+      "learning_rate": 2.6300347380766487e-06,
+      "loss": 0.1788,
+      "step": 50720
+    },
+    {
+      "epoch": 0.6677370111070169,
+      "grad_norm": 1.0710927248001099,
+      "learning_rate": 2.626282598402855e-06,
+      "loss": 0.1811,
+      "step": 50740
+    },
+    {
+      "epoch": 0.668000210559562,
+      "grad_norm": 1.5745491981506348,
+      "learning_rate": 2.6225321836113793e-06,
+      "loss": 0.1796,
+      "step": 50760
+    },
+    {
+      "epoch": 0.6682634100121072,
+      "grad_norm": 1.4582464694976807,
+      "learning_rate": 2.61878349642749e-06,
+      "loss": 0.1649,
+      "step": 50780
+    },
+    {
+      "epoch": 0.6685266094646524,
+      "grad_norm": 1.5733150243759155,
+      "learning_rate": 2.615036539575204e-06,
+      "loss": 0.1794,
+      "step": 50800
+    },
+    {
+      "epoch": 0.6687898089171974,
+      "grad_norm": 1.6788692474365234,
+      "learning_rate": 2.6112913157772813e-06,
+      "loss": 0.1891,
+      "step": 50820
+    },
+    {
+      "epoch": 0.6690530083697426,
+      "grad_norm": 1.2326730489730835,
+      "learning_rate": 2.6075478277552167e-06,
+      "loss": 0.1741,
+      "step": 50840
+    },
+    {
+      "epoch": 0.6693162078222877,
+      "grad_norm": 1.4938002824783325,
+      "learning_rate": 2.6038060782292556e-06,
+      "loss": 0.169,
+      "step": 50860
+    },
+    {
+      "epoch": 0.6695794072748329,
+      "grad_norm": 2.010051965713501,
+      "learning_rate": 2.6000660699183666e-06,
+      "loss": 0.1953,
+      "step": 50880
+    },
+    {
+      "epoch": 0.669842606727378,
+      "grad_norm": 1.6067085266113281,
+      "learning_rate": 2.596327805540259e-06,
+      "loss": 0.1699,
+      "step": 50900
+    },
+    {
+      "epoch": 0.6701058061799231,
+      "grad_norm": 1.474310278892517,
+      "learning_rate": 2.592591287811379e-06,
+      "loss": 0.1708,
+      "step": 50920
+    },
+    {
+      "epoch": 0.6703690056324683,
+      "grad_norm": 1.4701640605926514,
+      "learning_rate": 2.588856519446898e-06,
+      "loss": 0.191,
+      "step": 50940
+    },
+    {
+      "epoch": 0.6706322050850134,
+      "grad_norm": 1.7899872064590454,
+      "learning_rate": 2.5851235031607167e-06,
+      "loss": 0.1691,
+      "step": 50960
+    },
+    {
+      "epoch": 0.6708954045375586,
+      "grad_norm": 1.7956256866455078,
+      "learning_rate": 2.5813922416654656e-06,
+      "loss": 0.1661,
+      "step": 50980
+    },
+    {
+      "epoch": 0.6711586039901037,
+      "grad_norm": 1.4386402368545532,
+      "learning_rate": 2.5776627376724943e-06,
+      "loss": 0.1853,
+      "step": 51000
+    },
+    {
+      "epoch": 0.6711586039901037,
+      "eval_loss": 0.20490242540836334,
+      "eval_runtime": 4.3825,
+      "eval_samples_per_second": 3.651,
+      "eval_steps_per_second": 1.825,
+      "step": 51000
+    },
+    {
+      "epoch": 0.6714218034426488,
+      "grad_norm": 1.5740457773208618,
+      "learning_rate": 2.573934993891884e-06,
+      "loss": 0.189,
+      "step": 51020
+    },
+    {
+      "epoch": 0.671685002895194,
+      "grad_norm": 1.8175859451293945,
+      "learning_rate": 2.570209013032432e-06,
+      "loss": 0.1497,
+      "step": 51040
+    },
+    {
+      "epoch": 0.6719482023477391,
+      "grad_norm": 1.7222270965576172,
+      "learning_rate": 2.566484797801648e-06,
+      "loss": 0.1702,
+      "step": 51060
+    },
+    {
+      "epoch": 0.6722114018002843,
+      "grad_norm": 1.7358239889144897,
+      "learning_rate": 2.5627623509057733e-06,
+      "loss": 0.1721,
+      "step": 51080
+    },
+    {
+      "epoch": 0.6724746012528294,
+      "grad_norm": 1.6321637630462646,
+      "learning_rate": 2.5590416750497514e-06,
+      "loss": 0.1797,
+      "step": 51100
+    },
+    {
+      "epoch": 0.6727378007053745,
+      "grad_norm": 2.485515594482422,
+      "learning_rate": 2.5553227729372453e-06,
+      "loss": 0.1643,
+      "step": 51120
+    },
+    {
+      "epoch": 0.6730010001579196,
+      "grad_norm": 1.8552870750427246,
+      "learning_rate": 2.5516056472706267e-06,
+      "loss": 0.1794,
+      "step": 51140
+    },
+    {
+      "epoch": 0.6732641996104648,
+      "grad_norm": 1.8514764308929443,
+      "learning_rate": 2.5478903007509758e-06,
+      "loss": 0.1845,
+      "step": 51160
+    },
+    {
+      "epoch": 0.67352739906301,
+      "grad_norm": 1.6723122596740723,
+      "learning_rate": 2.5441767360780855e-06,
+      "loss": 0.1986,
+      "step": 51180
+    },
+    {
+      "epoch": 0.6737905985155551,
+      "grad_norm": 1.1502737998962402,
+      "learning_rate": 2.540464955950447e-06,
+      "loss": 0.1679,
+      "step": 51200
+    },
+    {
+      "epoch": 0.6740537979681003,
+      "grad_norm": 1.7347184419631958,
+      "learning_rate": 2.5367549630652578e-06,
+      "loss": 0.1687,
+      "step": 51220
+    },
+    {
+      "epoch": 0.6743169974206453,
+      "grad_norm": 1.6548160314559937,
+      "learning_rate": 2.533046760118416e-06,
+      "loss": 0.1675,
+      "step": 51240
+    },
+    {
+      "epoch": 0.6745801968731905,
+      "grad_norm": 1.6178314685821533,
+      "learning_rate": 2.5293403498045176e-06,
+      "loss": 0.187,
+      "step": 51260
+    },
+    {
+      "epoch": 0.6748433963257356,
+      "grad_norm": 1.729277491569519,
+      "learning_rate": 2.525635734816865e-06,
+      "loss": 0.1606,
+      "step": 51280
+    },
+    {
+      "epoch": 0.6751065957782808,
+      "grad_norm": 1.3853020668029785,
+      "learning_rate": 2.521932917847441e-06,
+      "loss": 0.1914,
+      "step": 51300
+    },
+    {
+      "epoch": 0.675369795230826,
+      "grad_norm": 1.449632167816162,
+      "learning_rate": 2.5182319015869306e-06,
+      "loss": 0.16,
+      "step": 51320
+    },
+    {
+      "epoch": 0.675632994683371,
+      "grad_norm": 1.8331776857376099,
+      "learning_rate": 2.5145326887247136e-06,
+      "loss": 0.18,
+      "step": 51340
+    },
+    {
+      "epoch": 0.6758961941359162,
+      "grad_norm": 1.9801384210586548,
+      "learning_rate": 2.510835281948851e-06,
+      "loss": 0.1636,
+      "step": 51360
+    },
+    {
+      "epoch": 0.6761593935884613,
+      "grad_norm": 2.209718704223633,
+      "learning_rate": 2.5071396839460983e-06,
+      "loss": 0.1622,
+      "step": 51380
+    },
+    {
+      "epoch": 0.6764225930410065,
+      "grad_norm": 2.225564956665039,
+      "learning_rate": 2.5034458974018925e-06,
+      "loss": 0.1668,
+      "step": 51400
+    },
+    {
+      "epoch": 0.6766857924935517,
+      "grad_norm": 1.5598043203353882,
+      "learning_rate": 2.4997539250003534e-06,
+      "loss": 0.1532,
+      "step": 51420
+    },
+    {
+      "epoch": 0.6769489919460967,
+      "grad_norm": 1.9549717903137207,
+      "learning_rate": 2.496063769424289e-06,
+      "loss": 0.1617,
+      "step": 51440
+    },
+    {
+      "epoch": 0.6772121913986419,
+      "grad_norm": 1.430174469947815,
+      "learning_rate": 2.492375433355181e-06,
+      "loss": 0.1677,
+      "step": 51460
+    },
+    {
+      "epoch": 0.677475390851187,
+      "grad_norm": 1.402801275253296,
+      "learning_rate": 2.48868891947319e-06,
+      "loss": 0.1831,
+      "step": 51480
+    },
+    {
+      "epoch": 0.6777385903037322,
+      "grad_norm": 1.9006798267364502,
+      "learning_rate": 2.485004230457154e-06,
+      "loss": 0.1549,
+      "step": 51500
+    },
+    {
+      "epoch": 0.6777385903037322,
+      "eval_loss": 0.20955519378185272,
+      "eval_runtime": 3.323,
+      "eval_samples_per_second": 4.815,
+      "eval_steps_per_second": 2.407,
+      "step": 51500
+    },
+    {
+      "epoch": 0.6780017897562773,
+      "grad_norm": 1.4688345193862915,
+      "learning_rate": 2.4813213689845846e-06,
+      "loss": 0.1766,
+      "step": 51520
+    },
+    {
+      "epoch": 0.6782649892088225,
+      "grad_norm": 1.4128401279449463,
+      "learning_rate": 2.477640337731664e-06,
+      "loss": 0.1642,
+      "step": 51540
+    },
+    {
+      "epoch": 0.6785281886613675,
+      "grad_norm": 1.668890118598938,
+      "learning_rate": 2.4739611393732456e-06,
+      "loss": 0.1796,
+      "step": 51560
+    },
+    {
+      "epoch": 0.6787913881139127,
+      "grad_norm": 1.2653101682662964,
+      "learning_rate": 2.470283776582849e-06,
+      "loss": 0.1843,
+      "step": 51580
+    },
+    {
+      "epoch": 0.6790545875664579,
+      "grad_norm": 1.8933205604553223,
+      "learning_rate": 2.4666082520326663e-06,
+      "loss": 0.1676,
+      "step": 51600
+    },
+    {
+      "epoch": 0.679317787019003,
+      "grad_norm": 1.5248860120773315,
+      "learning_rate": 2.462934568393546e-06,
+      "loss": 0.1913,
+      "step": 51620
+    },
+    {
+      "epoch": 0.6795809864715482,
+      "grad_norm": 1.6484791040420532,
+      "learning_rate": 2.4592627283350034e-06,
+      "loss": 0.1706,
+      "step": 51640
+    },
+    {
+      "epoch": 0.6798441859240932,
+      "grad_norm": 1.16077721118927,
+      "learning_rate": 2.4555927345252125e-06,
+      "loss": 0.1818,
+      "step": 51660
+    },
+    {
+      "epoch": 0.6801073853766384,
+      "grad_norm": 1.411294937133789,
+      "learning_rate": 2.451924589631004e-06,
+      "loss": 0.1442,
+      "step": 51680
+    },
+    {
+      "epoch": 0.6803705848291836,
+      "grad_norm": 1.653726577758789,
+      "learning_rate": 2.4482582963178727e-06,
+      "loss": 0.1809,
+      "step": 51700
+    },
+    {
+      "epoch": 0.6806337842817287,
+      "grad_norm": 2.0255932807922363,
+      "learning_rate": 2.4445938572499543e-06,
+      "loss": 0.1726,
+      "step": 51720
+    },
+    {
+      "epoch": 0.6808969837342739,
+      "grad_norm": 1.4002834558486938,
+      "learning_rate": 2.4409312750900526e-06,
+      "loss": 0.1699,
+      "step": 51740
+    },
+    {
+      "epoch": 0.6811601831868189,
+      "grad_norm": 1.7429333925247192,
+      "learning_rate": 2.4372705524996116e-06,
+      "loss": 0.1641,
+      "step": 51760
+    },
+    {
+      "epoch": 0.6814233826393641,
+      "grad_norm": 1.363295316696167,
+      "learning_rate": 2.433611692138727e-06,
+      "loss": 0.1744,
+      "step": 51780
+    },
+    {
+      "epoch": 0.6816865820919092,
+      "grad_norm": 1.7755811214447021,
+      "learning_rate": 2.429954696666142e-06,
+      "loss": 0.1739,
+      "step": 51800
+    },
+    {
+      "epoch": 0.6819497815444544,
+      "grad_norm": 1.8343369960784912,
+      "learning_rate": 2.426299568739243e-06,
+      "loss": 0.1738,
+      "step": 51820
+    },
+    {
+      "epoch": 0.6822129809969996,
+      "grad_norm": 1.218808650970459,
+      "learning_rate": 2.4226463110140598e-06,
+      "loss": 0.1837,
+      "step": 51840
+    },
+    {
+      "epoch": 0.6824761804495446,
+      "grad_norm": 2.016817331314087,
+      "learning_rate": 2.4189949261452688e-06,
+      "loss": 0.1778,
+      "step": 51860
+    },
+    {
+      "epoch": 0.6827393799020898,
+      "grad_norm": 1.8033109903335571,
+      "learning_rate": 2.4153454167861728e-06,
+      "loss": 0.1676,
+      "step": 51880
+    },
+    {
+      "epoch": 0.6830025793546349,
+      "grad_norm": 1.6130977869033813,
+      "learning_rate": 2.411697785588725e-06,
+      "loss": 0.1823,
+      "step": 51900
+    },
+    {
+      "epoch": 0.6832657788071801,
+      "grad_norm": 1.627455234527588,
+      "learning_rate": 2.4080520352035065e-06,
+      "loss": 0.1759,
+      "step": 51920
+    },
+    {
+      "epoch": 0.6835289782597253,
+      "grad_norm": 2.137490749359131,
+      "learning_rate": 2.4044081682797337e-06,
+      "loss": 0.1682,
+      "step": 51940
+    },
+    {
+      "epoch": 0.6837921777122704,
+      "grad_norm": 1.6552345752716064,
+      "learning_rate": 2.400766187465254e-06,
+      "loss": 0.1751,
+      "step": 51960
+    },
+    {
+      "epoch": 0.6840553771648155,
+      "grad_norm": 1.843685507774353,
+      "learning_rate": 2.3971260954065413e-06,
+      "loss": 0.191,
+      "step": 51980
+    },
+    {
+      "epoch": 0.6843185766173606,
+      "grad_norm": 1.2494618892669678,
+      "learning_rate": 2.3934878947487055e-06,
+      "loss": 0.19,
+      "step": 52000
+    },
+    {
+      "epoch": 0.6843185766173606,
+      "eval_loss": 0.20941147208213806,
+      "eval_runtime": 4.4407,
+      "eval_samples_per_second": 3.603,
+      "eval_steps_per_second": 1.802,
+      "step": 52000
+    },
+    {
+      "epoch": 0.6845817760699058,
+      "grad_norm": 1.0327956676483154,
+      "learning_rate": 2.3898515881354735e-06,
+      "loss": 0.1768,
+      "step": 52020
+    },
+    {
+      "epoch": 0.6848449755224509,
+      "grad_norm": 1.5938191413879395,
+      "learning_rate": 2.3862171782091997e-06,
+      "loss": 0.195,
+      "step": 52040
+    },
+    {
+      "epoch": 0.6851081749749961,
+      "grad_norm": 1.3577386140823364,
+      "learning_rate": 2.3825846676108592e-06,
+      "loss": 0.1759,
+      "step": 52060
+    },
+    {
+      "epoch": 0.6853713744275411,
+      "grad_norm": 1.6651602983474731,
+      "learning_rate": 2.3789540589800476e-06,
+      "loss": 0.1756,
+      "step": 52080
+    },
+    {
+      "epoch": 0.6856345738800863,
+      "grad_norm": 1.2678914070129395,
+      "learning_rate": 2.3753253549549778e-06,
+      "loss": 0.1713,
+      "step": 52100
+    },
+    {
+      "epoch": 0.6858977733326315,
+      "grad_norm": 1.3825703859329224,
+      "learning_rate": 2.371698558172479e-06,
+      "loss": 0.1583,
+      "step": 52120
+    },
+    {
+      "epoch": 0.6861609727851766,
+      "grad_norm": 1.7462259531021118,
+      "learning_rate": 2.368073671267993e-06,
+      "loss": 0.1809,
+      "step": 52140
+    },
+    {
+      "epoch": 0.6864241722377218,
+      "grad_norm": 1.8407714366912842,
+      "learning_rate": 2.3644506968755787e-06,
+      "loss": 0.1614,
+      "step": 52160
+    },
+    {
+      "epoch": 0.6866873716902668,
+      "grad_norm": 1.370564341545105,
+      "learning_rate": 2.3608296376278998e-06,
+      "loss": 0.1827,
+      "step": 52180
+    },
+    {
+      "epoch": 0.686950571142812,
+      "grad_norm": 1.8382662534713745,
+      "learning_rate": 2.357210496156231e-06,
+      "loss": 0.1652,
+      "step": 52200
+    },
+    {
+      "epoch": 0.6872137705953572,
+      "grad_norm": 1.5238745212554932,
+      "learning_rate": 2.3535932750904526e-06,
+      "loss": 0.1866,
+      "step": 52220
+    },
+    {
+      "epoch": 0.6874769700479023,
+      "grad_norm": 1.8697549104690552,
+      "learning_rate": 2.3499779770590475e-06,
+      "loss": 0.1576,
+      "step": 52240
+    },
+    {
+      "epoch": 0.6877401695004475,
+      "grad_norm": 1.2097283601760864,
+      "learning_rate": 2.346364604689107e-06,
+      "loss": 0.1801,
+      "step": 52260
+    },
+    {
+      "epoch": 0.6880033689529925,
+      "grad_norm": 1.500579595565796,
+      "learning_rate": 2.34275316060632e-06,
+      "loss": 0.1651,
+      "step": 52280
+    },
+    {
+      "epoch": 0.6882665684055377,
+      "grad_norm": 1.5092509984970093,
+      "learning_rate": 2.3391436474349673e-06,
+      "loss": 0.1761,
+      "step": 52300
+    },
+    {
+      "epoch": 0.6885297678580828,
+      "grad_norm": 1.1442841291427612,
+      "learning_rate": 2.3355360677979385e-06,
+      "loss": 0.191,
+      "step": 52320
+    },
+    {
+      "epoch": 0.688792967310628,
+      "grad_norm": 2.0138142108917236,
+      "learning_rate": 2.3319304243167097e-06,
+      "loss": 0.16,
+      "step": 52340
+    },
+    {
+      "epoch": 0.6890561667631732,
+      "grad_norm": 1.681333065032959,
+      "learning_rate": 2.3283267196113577e-06,
+      "loss": 0.1876,
+      "step": 52360
+    },
+    {
+      "epoch": 0.6893193662157183,
+      "grad_norm": 1.6829001903533936,
+      "learning_rate": 2.3247249563005415e-06,
+      "loss": 0.1703,
+      "step": 52380
+    },
+    {
+      "epoch": 0.6895825656682634,
+      "grad_norm": 1.9380804300308228,
+      "learning_rate": 2.321125137001512e-06,
+      "loss": 0.1813,
+      "step": 52400
+    },
+    {
+      "epoch": 0.6898457651208085,
+      "grad_norm": 1.1389501094818115,
+      "learning_rate": 2.3175272643301134e-06,
+      "loss": 0.1824,
+      "step": 52420
+    },
+    {
+      "epoch": 0.6901089645733537,
+      "grad_norm": 1.4777799844741821,
+      "learning_rate": 2.3139313409007697e-06,
+      "loss": 0.1758,
+      "step": 52440
+    },
+    {
+      "epoch": 0.6903721640258988,
+      "grad_norm": 1.7042018175125122,
+      "learning_rate": 2.310337369326489e-06,
+      "loss": 0.1894,
+      "step": 52460
+    },
+    {
+      "epoch": 0.690635363478444,
+      "grad_norm": 2.0535926818847656,
+      "learning_rate": 2.306745352218862e-06,
+      "loss": 0.1775,
+      "step": 52480
+    },
+    {
+      "epoch": 0.6908985629309891,
+      "grad_norm": 1.590975284576416,
+      "learning_rate": 2.3031552921880577e-06,
+      "loss": 0.1894,
+      "step": 52500
+    },
+    {
+      "epoch": 0.6908985629309891,
+      "eval_loss": 0.20782417058944702,
+      "eval_runtime": 4.319,
+      "eval_samples_per_second": 3.705,
+      "eval_steps_per_second": 1.852,
+      "step": 52500
+    },
+    {
+      "epoch": 0.6911617623835342,
+      "grad_norm": 1.681665301322937,
+      "learning_rate": 2.2995671918428305e-06,
+      "loss": 0.1781,
+      "step": 52520
+    },
+    {
+      "epoch": 0.6914249618360794,
+      "grad_norm": 1.7440110445022583,
+      "learning_rate": 2.2959810537904976e-06,
+      "loss": 0.1784,
+      "step": 52540
+    },
+    {
+      "epoch": 0.6916881612886245,
+      "grad_norm": 1.2990626096725464,
+      "learning_rate": 2.2923968806369573e-06,
+      "loss": 0.1785,
+      "step": 52560
+    },
+    {
+      "epoch": 0.6919513607411697,
+      "grad_norm": 1.8547981977462769,
+      "learning_rate": 2.288814674986685e-06,
+      "loss": 0.1854,
+      "step": 52580
+    },
+    {
+      "epoch": 0.6922145601937147,
+      "grad_norm": 1.609271764755249,
+      "learning_rate": 2.285234439442719e-06,
+      "loss": 0.187,
+      "step": 52600
+    },
+    {
+      "epoch": 0.6924777596462599,
+      "grad_norm": 1.402474045753479,
+      "learning_rate": 2.2816561766066682e-06,
+      "loss": 0.1861,
+      "step": 52620
+    },
+    {
+      "epoch": 0.6927409590988051,
+      "grad_norm": 2.1044745445251465,
+      "learning_rate": 2.278079889078709e-06,
+      "loss": 0.1781,
+      "step": 52640
+    },
+    {
+      "epoch": 0.6930041585513502,
+      "grad_norm": 1.615842342376709,
+      "learning_rate": 2.2745055794575786e-06,
+      "loss": 0.1754,
+      "step": 52660
+    },
+    {
+      "epoch": 0.6932673580038954,
+      "grad_norm": 1.2315571308135986,
+      "learning_rate": 2.270933250340585e-06,
+      "loss": 0.1707,
+      "step": 52680
+    },
+    {
+      "epoch": 0.6935305574564405,
+      "grad_norm": 1.1345582008361816,
+      "learning_rate": 2.2673629043235918e-06,
+      "loss": 0.176,
+      "step": 52700
+    },
+    {
+      "epoch": 0.6937937569089856,
+      "grad_norm": 2.0543243885040283,
+      "learning_rate": 2.2637945440010164e-06,
+      "loss": 0.1519,
+      "step": 52720
+    },
+    {
+      "epoch": 0.6940569563615308,
+      "grad_norm": 1.8855398893356323,
+      "learning_rate": 2.2602281719658435e-06,
+      "loss": 0.1852,
+      "step": 52740
+    },
+    {
+      "epoch": 0.6943201558140759,
+      "grad_norm": 1.6052981615066528,
+      "learning_rate": 2.2566637908096073e-06,
+      "loss": 0.1794,
+      "step": 52760
+    },
+    {
+      "epoch": 0.6945833552666211,
+      "grad_norm": 1.9944682121276855,
+      "learning_rate": 2.2531014031223965e-06,
+      "loss": 0.1714,
+      "step": 52780
+    },
+    {
+      "epoch": 0.6948465547191662,
+      "grad_norm": 1.178808569908142,
+      "learning_rate": 2.2495410114928495e-06,
+      "loss": 0.1649,
+      "step": 52800
+    },
+    {
+      "epoch": 0.6951097541717113,
+      "grad_norm": 1.8873926401138306,
+      "learning_rate": 2.245982618508155e-06,
+      "loss": 0.1751,
+      "step": 52820
+    },
+    {
+      "epoch": 0.6953729536242564,
+      "grad_norm": 1.5615321397781372,
+      "learning_rate": 2.2424262267540537e-06,
+      "loss": 0.1624,
+      "step": 52840
+    },
+    {
+      "epoch": 0.6956361530768016,
+      "grad_norm": 2.2063608169555664,
+      "learning_rate": 2.2388718388148266e-06,
+      "loss": 0.1653,
+      "step": 52860
+    },
+    {
+      "epoch": 0.6958993525293468,
+      "grad_norm": 2.025057792663574,
+      "learning_rate": 2.2353194572733004e-06,
+      "loss": 0.1753,
+      "step": 52880
+    },
+    {
+      "epoch": 0.6961625519818919,
+      "grad_norm": 1.4953291416168213,
+      "learning_rate": 2.2317690847108436e-06,
+      "loss": 0.1727,
+      "step": 52900
+    },
+    {
+      "epoch": 0.696425751434437,
+      "grad_norm": 1.9743825197219849,
+      "learning_rate": 2.2282207237073644e-06,
+      "loss": 0.165,
+      "step": 52920
+    },
+    {
+      "epoch": 0.6966889508869821,
+      "grad_norm": 1.6791150569915771,
+      "learning_rate": 2.224674376841314e-06,
+      "loss": 0.1799,
+      "step": 52940
+    },
+    {
+      "epoch": 0.6969521503395273,
+      "grad_norm": 1.513907551765442,
+      "learning_rate": 2.2211300466896717e-06,
+      "loss": 0.1664,
+      "step": 52960
+    },
+    {
+      "epoch": 0.6972153497920724,
+      "grad_norm": 1.8856406211853027,
+      "learning_rate": 2.217587735827956e-06,
+      "loss": 0.1693,
+      "step": 52980
+    },
+    {
+      "epoch": 0.6974785492446176,
+      "grad_norm": 1.9564813375473022,
+      "learning_rate": 2.2140474468302196e-06,
+      "loss": 0.1953,
+      "step": 53000
+    },
+    {
+      "epoch": 0.6974785492446176,
+      "eval_loss": 0.20792515575885773,
+      "eval_runtime": 4.3598,
+      "eval_samples_per_second": 3.67,
+      "eval_steps_per_second": 1.835,
+      "step": 53000
+    },
+    {
+      "epoch": 0.6977417486971627,
+      "grad_norm": 1.537707805633545,
+      "learning_rate": 2.210509182269045e-06,
+      "loss": 0.1732,
+      "step": 53020
+    },
+    {
+      "epoch": 0.6980049481497078,
+      "grad_norm": 1.7441264390945435,
+      "learning_rate": 2.2069729447155407e-06,
+      "loss": 0.1861,
+      "step": 53040
+    },
+    {
+      "epoch": 0.698268147602253,
+      "grad_norm": 1.5876779556274414,
+      "learning_rate": 2.203438736739345e-06,
+      "loss": 0.1581,
+      "step": 53060
+    },
+    {
+      "epoch": 0.6985313470547981,
+      "grad_norm": 1.7468292713165283,
+      "learning_rate": 2.199906560908619e-06,
+      "loss": 0.1849,
+      "step": 53080
+    },
+    {
+      "epoch": 0.6987945465073433,
+      "grad_norm": 1.662034511566162,
+      "learning_rate": 2.1963764197900526e-06,
+      "loss": 0.1679,
+      "step": 53100
+    },
+    {
+      "epoch": 0.6990577459598885,
+      "grad_norm": 1.5966877937316895,
+      "learning_rate": 2.192848315948852e-06,
+      "loss": 0.1728,
+      "step": 53120
+    },
+    {
+      "epoch": 0.6993209454124335,
+      "grad_norm": 1.69669771194458,
+      "learning_rate": 2.1893222519487433e-06,
+      "loss": 0.1731,
+      "step": 53140
+    },
+    {
+      "epoch": 0.6995841448649787,
+      "grad_norm": 2.5742835998535156,
+      "learning_rate": 2.185798230351972e-06,
+      "loss": 0.1733,
+      "step": 53160
+    },
+    {
+      "epoch": 0.6998473443175238,
+      "grad_norm": 1.7248865365982056,
+      "learning_rate": 2.1822762537193e-06,
+      "loss": 0.1742,
+      "step": 53180
+    },
+    {
+      "epoch": 0.700110543770069,
+      "grad_norm": 1.7163605690002441,
+      "learning_rate": 2.1787563246100003e-06,
+      "loss": 0.1695,
+      "step": 53200
+    },
+    {
+      "epoch": 0.7003737432226141,
+      "grad_norm": 1.6984111070632935,
+      "learning_rate": 2.1752384455818608e-06,
+      "loss": 0.1663,
+      "step": 53220
+    },
+    {
+      "epoch": 0.7006369426751592,
+      "grad_norm": 1.911356806755066,
+      "learning_rate": 2.1717226191911765e-06,
+      "loss": 0.1866,
+      "step": 53240
+    },
+    {
+      "epoch": 0.7009001421277044,
+      "grad_norm": 1.178343415260315,
+      "learning_rate": 2.168208847992758e-06,
+      "loss": 0.1684,
+      "step": 53260
+    },
+    {
+      "epoch": 0.7011633415802495,
+      "grad_norm": 1.551300287246704,
+      "learning_rate": 2.1646971345399142e-06,
+      "loss": 0.1606,
+      "step": 53280
+    },
+    {
+      "epoch": 0.7014265410327947,
+      "grad_norm": 1.2357263565063477,
+      "learning_rate": 2.1611874813844636e-06,
+      "loss": 0.166,
+      "step": 53300
+    },
+    {
+      "epoch": 0.7016897404853398,
+      "grad_norm": 1.5292786359786987,
+      "learning_rate": 2.1576798910767248e-06,
+      "loss": 0.1707,
+      "step": 53320
+    },
+    {
+      "epoch": 0.7019529399378849,
+      "grad_norm": 1.4368942975997925,
+      "learning_rate": 2.1541743661655175e-06,
+      "loss": 0.1643,
+      "step": 53340
+    },
+    {
+      "epoch": 0.70221613939043,
+      "grad_norm": 1.6769884824752808,
+      "learning_rate": 2.150670909198167e-06,
+      "loss": 0.167,
+      "step": 53360
+    },
+    {
+      "epoch": 0.7024793388429752,
+      "grad_norm": 1.6338958740234375,
+      "learning_rate": 2.147169522720483e-06,
+      "loss": 0.1772,
+      "step": 53380
+    },
+    {
+      "epoch": 0.7027425382955204,
+      "grad_norm": 2.3556907176971436,
+      "learning_rate": 2.1436702092767847e-06,
+      "loss": 0.1735,
+      "step": 53400
+    },
+    {
+      "epoch": 0.7030057377480655,
+      "grad_norm": 1.5587173700332642,
+      "learning_rate": 2.140172971409876e-06,
+      "loss": 0.1573,
+      "step": 53420
+    },
+    {
+      "epoch": 0.7032689372006107,
+      "grad_norm": 2.361372709274292,
+      "learning_rate": 2.1366778116610557e-06,
+      "loss": 0.1809,
+      "step": 53440
+    },
+    {
+      "epoch": 0.7035321366531557,
+      "grad_norm": 1.5027693510055542,
+      "learning_rate": 2.1331847325701115e-06,
+      "loss": 0.1675,
+      "step": 53460
+    },
+    {
+      "epoch": 0.7037953361057009,
+      "grad_norm": 1.6620776653289795,
+      "learning_rate": 2.1296937366753197e-06,
+      "loss": 0.1797,
+      "step": 53480
+    },
+    {
+      "epoch": 0.704058535558246,
+      "grad_norm": 1.3067456483840942,
+      "learning_rate": 2.12620482651344e-06,
+      "loss": 0.1614,
+      "step": 53500
+    },
+    {
+      "epoch": 0.704058535558246,
+      "eval_loss": 0.20561540126800537,
+      "eval_runtime": 4.2972,
+      "eval_samples_per_second": 3.723,
+      "eval_steps_per_second": 1.862,
+      "step": 53500
+    },
+    {
+      "epoch": 0.7043217350107912,
+      "grad_norm": 1.9893851280212402,
+      "learning_rate": 2.122718004619726e-06,
+      "loss": 0.1669,
+      "step": 53520
+    },
+    {
+      "epoch": 0.7045849344633364,
+      "grad_norm": 1.8841345310211182,
+      "learning_rate": 2.1192332735278996e-06,
+      "loss": 0.1693,
+      "step": 53540
+    },
+    {
+      "epoch": 0.7048481339158814,
+      "grad_norm": 1.4480974674224854,
+      "learning_rate": 2.1157506357701756e-06,
+      "loss": 0.1801,
+      "step": 53560
+    },
+    {
+      "epoch": 0.7051113333684266,
+      "grad_norm": 1.5668474435806274,
+      "learning_rate": 2.112270093877241e-06,
+      "loss": 0.1766,
+      "step": 53580
+    },
+    {
+      "epoch": 0.7053745328209717,
+      "grad_norm": 1.2195556163787842,
+      "learning_rate": 2.108791650378263e-06,
+      "loss": 0.1629,
+      "step": 53600
+    },
+    {
+      "epoch": 0.7056377322735169,
+      "grad_norm": 1.4103891849517822,
+      "learning_rate": 2.1053153078008816e-06,
+      "loss": 0.1711,
+      "step": 53620
+    },
+    {
+      "epoch": 0.705900931726062,
+      "grad_norm": 1.225401520729065,
+      "learning_rate": 2.1018410686712095e-06,
+      "loss": 0.171,
+      "step": 53640
+    },
+    {
+      "epoch": 0.7061641311786071,
+      "grad_norm": 1.722366452217102,
+      "learning_rate": 2.0983689355138366e-06,
+      "loss": 0.1758,
+      "step": 53660
+    },
+    {
+      "epoch": 0.7064273306311523,
+      "grad_norm": 1.7433558702468872,
+      "learning_rate": 2.0948989108518163e-06,
+      "loss": 0.1819,
+      "step": 53680
+    },
+    {
+      "epoch": 0.7066905300836974,
+      "grad_norm": 1.6438217163085938,
+      "learning_rate": 2.0914309972066728e-06,
+      "loss": 0.1728,
+      "step": 53700
+    },
+    {
+      "epoch": 0.7069537295362426,
+      "grad_norm": 1.836316704750061,
+      "learning_rate": 2.0879651970983948e-06,
+      "loss": 0.1615,
+      "step": 53720
+    },
+    {
+      "epoch": 0.7072169289887877,
+      "grad_norm": 1.409497857093811,
+      "learning_rate": 2.084501513045435e-06,
+      "loss": 0.1687,
+      "step": 53740
+    },
+    {
+      "epoch": 0.7074801284413328,
+      "grad_norm": 1.4985322952270508,
+      "learning_rate": 2.0810399475647137e-06,
+      "loss": 0.1934,
+      "step": 53760
+    },
+    {
+      "epoch": 0.707743327893878,
+      "grad_norm": 1.4153122901916504,
+      "learning_rate": 2.0775805031716033e-06,
+      "loss": 0.1775,
+      "step": 53780
+    },
+    {
+      "epoch": 0.7080065273464231,
+      "grad_norm": 1.405444622039795,
+      "learning_rate": 2.0741231823799385e-06,
+      "loss": 0.1701,
+      "step": 53800
+    },
+    {
+      "epoch": 0.7082697267989683,
+      "grad_norm": 1.9892512559890747,
+      "learning_rate": 2.0706679877020145e-06,
+      "loss": 0.1745,
+      "step": 53820
+    },
+    {
+      "epoch": 0.7085329262515134,
+      "grad_norm": 1.460495948791504,
+      "learning_rate": 2.0672149216485775e-06,
+      "loss": 0.1591,
+      "step": 53840
+    },
+    {
+      "epoch": 0.7087961257040586,
+      "grad_norm": 1.8487451076507568,
+      "learning_rate": 2.063763986728828e-06,
+      "loss": 0.1699,
+      "step": 53860
+    },
+    {
+      "epoch": 0.7090593251566036,
+      "grad_norm": 1.6382486820220947,
+      "learning_rate": 2.0603151854504165e-06,
+      "loss": 0.1726,
+      "step": 53880
+    },
+    {
+      "epoch": 0.7093225246091488,
+      "grad_norm": 1.980622410774231,
+      "learning_rate": 2.056868520319443e-06,
+      "loss": 0.1669,
+      "step": 53900
+    },
+    {
+      "epoch": 0.709585724061694,
+      "grad_norm": 1.8458257913589478,
+      "learning_rate": 2.0534239938404615e-06,
+      "loss": 0.166,
+      "step": 53920
+    },
+    {
+      "epoch": 0.7098489235142391,
+      "grad_norm": 1.731569528579712,
+      "learning_rate": 2.0499816085164653e-06,
+      "loss": 0.1874,
+      "step": 53940
+    },
+    {
+      "epoch": 0.7101121229667843,
+      "grad_norm": 1.7408175468444824,
+      "learning_rate": 2.0465413668488894e-06,
+      "loss": 0.1696,
+      "step": 53960
+    },
+    {
+      "epoch": 0.7103753224193293,
+      "grad_norm": 1.3498642444610596,
+      "learning_rate": 2.0431032713376194e-06,
+      "loss": 0.1728,
+      "step": 53980
+    },
+    {
+      "epoch": 0.7106385218718745,
+      "grad_norm": 1.4871430397033691,
+      "learning_rate": 2.0396673244809743e-06,
+      "loss": 0.1909,
+      "step": 54000
+    },
+    {
+      "epoch": 0.7106385218718745,
+      "eval_loss": 0.20290595293045044,
+      "eval_runtime": 4.2127,
+      "eval_samples_per_second": 3.798,
+      "eval_steps_per_second": 1.899,
+      "step": 54000
+    },
+    {
+      "epoch": 0.7109017213244196,
+      "grad_norm": 1.9374583959579468,
+      "learning_rate": 2.03623352877572e-06,
+      "loss": 0.1638,
+      "step": 54020
+    },
+    {
+      "epoch": 0.7111649207769648,
+      "grad_norm": 1.5008760690689087,
+      "learning_rate": 2.0328018867170487e-06,
+      "loss": 0.1898,
+      "step": 54040
+    },
+    {
+      "epoch": 0.71142812022951,
+      "grad_norm": 1.720381498336792,
+      "learning_rate": 2.0293724007985922e-06,
+      "loss": 0.1753,
+      "step": 54060
+    },
+    {
+      "epoch": 0.711691319682055,
+      "grad_norm": 1.8450144529342651,
+      "learning_rate": 2.0259450735124205e-06,
+      "loss": 0.1605,
+      "step": 54080
+    },
+    {
+      "epoch": 0.7119545191346002,
+      "grad_norm": 1.2642452716827393,
+      "learning_rate": 2.022519907349028e-06,
+      "loss": 0.1634,
+      "step": 54100
+    },
+    {
+      "epoch": 0.7122177185871453,
+      "grad_norm": 1.2330777645111084,
+      "learning_rate": 2.0190969047973422e-06,
+      "loss": 0.1891,
+      "step": 54120
+    },
+    {
+      "epoch": 0.7124809180396905,
+      "grad_norm": 1.4682245254516602,
+      "learning_rate": 2.0156760683447175e-06,
+      "loss": 0.1724,
+      "step": 54140
+    },
+    {
+      "epoch": 0.7127441174922357,
+      "grad_norm": 1.3385999202728271,
+      "learning_rate": 2.012257400476932e-06,
+      "loss": 0.1853,
+      "step": 54160
+    },
+    {
+      "epoch": 0.7130073169447807,
+      "grad_norm": 1.2389012575149536,
+      "learning_rate": 2.0088409036781964e-06,
+      "loss": 0.189,
+      "step": 54180
+    },
+    {
+      "epoch": 0.7132705163973259,
+      "grad_norm": 2.013472080230713,
+      "learning_rate": 2.005426580431131e-06,
+      "loss": 0.1779,
+      "step": 54200
+    },
+    {
+      "epoch": 0.713533715849871,
+      "grad_norm": 1.7019222974777222,
+      "learning_rate": 2.002014433216785e-06,
+      "loss": 0.1968,
+      "step": 54220
+    },
+    {
+      "epoch": 0.7137969153024162,
+      "grad_norm": 1.5610332489013672,
+      "learning_rate": 1.998604464514628e-06,
+      "loss": 0.1875,
+      "step": 54240
+    },
+    {
+      "epoch": 0.7140601147549613,
+      "grad_norm": 1.734917402267456,
+      "learning_rate": 1.9951966768025406e-06,
+      "loss": 0.1761,
+      "step": 54260
+    },
+    {
+      "epoch": 0.7143233142075065,
+      "grad_norm": 1.0353838205337524,
+      "learning_rate": 1.9917910725568216e-06,
+      "loss": 0.1693,
+      "step": 54280
+    },
+    {
+      "epoch": 0.7145865136600515,
+      "grad_norm": 1.4750676155090332,
+      "learning_rate": 1.9883876542521835e-06,
+      "loss": 0.1587,
+      "step": 54300
+    },
+    {
+      "epoch": 0.7148497131125967,
+      "grad_norm": 1.6739774942398071,
+      "learning_rate": 1.9849864243617465e-06,
+      "loss": 0.2121,
+      "step": 54320
+    },
+    {
+      "epoch": 0.7151129125651419,
+      "grad_norm": 1.9324262142181396,
+      "learning_rate": 1.9815873853570495e-06,
+      "loss": 0.1753,
+      "step": 54340
+    },
+    {
+      "epoch": 0.715376112017687,
+      "grad_norm": 1.5063908100128174,
+      "learning_rate": 1.9781905397080315e-06,
+      "loss": 0.1692,
+      "step": 54360
+    },
+    {
+      "epoch": 0.7156393114702322,
+      "grad_norm": 1.5037752389907837,
+      "learning_rate": 1.974795889883036e-06,
+      "loss": 0.1733,
+      "step": 54380
+    },
+    {
+      "epoch": 0.7159025109227772,
+      "grad_norm": 1.7726136445999146,
+      "learning_rate": 1.971403438348819e-06,
+      "loss": 0.1554,
+      "step": 54400
+    },
+    {
+      "epoch": 0.7161657103753224,
+      "grad_norm": 1.503907561302185,
+      "learning_rate": 1.968013187570534e-06,
+      "loss": 0.1892,
+      "step": 54420
+    },
+    {
+      "epoch": 0.7164289098278676,
+      "grad_norm": 1.5476974248886108,
+      "learning_rate": 1.964625140011735e-06,
+      "loss": 0.169,
+      "step": 54440
+    },
+    {
+      "epoch": 0.7166921092804127,
+      "grad_norm": 1.7398818731307983,
+      "learning_rate": 1.961239298134377e-06,
+      "loss": 0.1783,
+      "step": 54460
+    },
+    {
+      "epoch": 0.7169553087329579,
+      "grad_norm": 1.421011209487915,
+      "learning_rate": 1.957855664398809e-06,
+      "loss": 0.1782,
+      "step": 54480
+    },
+    {
+      "epoch": 0.7172185081855029,
+      "grad_norm": 2.0697684288024902,
+      "learning_rate": 1.954474241263782e-06,
+      "loss": 0.1669,
+      "step": 54500
+    },
+    {
+      "epoch": 0.7172185081855029,
+      "eval_loss": 0.2050013542175293,
+      "eval_runtime": 4.6996,
+      "eval_samples_per_second": 3.405,
+      "eval_steps_per_second": 1.702,
+      "step": 54500
+    },
+    {
+      "epoch": 0.7174817076380481,
+      "grad_norm": 1.5358651876449585,
+      "learning_rate": 1.9510950311864333e-06,
+      "loss": 0.1612,
+      "step": 54520
+    },
+    {
+      "epoch": 0.7177449070905932,
+      "grad_norm": 1.3975154161453247,
+      "learning_rate": 1.947718036622297e-06,
+      "loss": 0.1733,
+      "step": 54540
+    },
+    {
+      "epoch": 0.7180081065431384,
+      "grad_norm": 1.739821434020996,
+      "learning_rate": 1.9443432600252936e-06,
+      "loss": 0.1965,
+      "step": 54560
+    },
+    {
+      "epoch": 0.7182713059956836,
+      "grad_norm": 1.9556552171707153,
+      "learning_rate": 1.9409707038477337e-06,
+      "loss": 0.1897,
+      "step": 54580
+    },
+    {
+      "epoch": 0.7185345054482287,
+      "grad_norm": 1.5513521432876587,
+      "learning_rate": 1.937600370540319e-06,
+      "loss": 0.1892,
+      "step": 54600
+    },
+    {
+      "epoch": 0.7187977049007738,
+      "grad_norm": 1.5623219013214111,
+      "learning_rate": 1.9342322625521253e-06,
+      "loss": 0.1724,
+      "step": 54620
+    },
+    {
+      "epoch": 0.7190609043533189,
+      "grad_norm": 1.8317244052886963,
+      "learning_rate": 1.930866382330619e-06,
+      "loss": 0.1764,
+      "step": 54640
+    },
+    {
+      "epoch": 0.7193241038058641,
+      "grad_norm": 1.8606574535369873,
+      "learning_rate": 1.927502732321649e-06,
+      "loss": 0.1787,
+      "step": 54660
+    },
+    {
+      "epoch": 0.7195873032584092,
+      "grad_norm": 1.975109577178955,
+      "learning_rate": 1.924141314969438e-06,
+      "loss": 0.1733,
+      "step": 54680
+    },
+    {
+      "epoch": 0.7198505027109544,
+      "grad_norm": 1.551741123199463,
+      "learning_rate": 1.920782132716591e-06,
+      "loss": 0.1708,
+      "step": 54700
+    },
+    {
+      "epoch": 0.7201137021634995,
+      "grad_norm": 1.2717106342315674,
+      "learning_rate": 1.917425188004085e-06,
+      "loss": 0.1852,
+      "step": 54720
+    },
+    {
+      "epoch": 0.7203769016160446,
+      "grad_norm": 1.925411343574524,
+      "learning_rate": 1.914070483271272e-06,
+      "loss": 0.1905,
+      "step": 54740
+    },
+    {
+      "epoch": 0.7206401010685898,
+      "grad_norm": 1.55620276927948,
+      "learning_rate": 1.910718020955883e-06,
+      "loss": 0.1736,
+      "step": 54760
+    },
+    {
+      "epoch": 0.7209033005211349,
+      "grad_norm": 1.0281237363815308,
+      "learning_rate": 1.9073678034940056e-06,
+      "loss": 0.179,
+      "step": 54780
+    },
+    {
+      "epoch": 0.7211664999736801,
+      "grad_norm": 1.7794556617736816,
+      "learning_rate": 1.904019833320111e-06,
+      "loss": 0.1733,
+      "step": 54800
+    },
+    {
+      "epoch": 0.7214296994262251,
+      "grad_norm": 2.8484020233154297,
+      "learning_rate": 1.9006741128670276e-06,
+      "loss": 0.1654,
+      "step": 54820
+    },
+    {
+      "epoch": 0.7216928988787703,
+      "grad_norm": 1.5727473497390747,
+      "learning_rate": 1.8973306445659533e-06,
+      "loss": 0.167,
+      "step": 54840
+    },
+    {
+      "epoch": 0.7219560983313155,
+      "grad_norm": 1.4796510934829712,
+      "learning_rate": 1.8939894308464474e-06,
+      "loss": 0.1808,
+      "step": 54860
+    },
+    {
+      "epoch": 0.7222192977838606,
+      "grad_norm": 1.4627619981765747,
+      "learning_rate": 1.890650474136433e-06,
+      "loss": 0.1755,
+      "step": 54880
+    },
+    {
+      "epoch": 0.7224824972364058,
+      "grad_norm": 1.519425392150879,
+      "learning_rate": 1.8873137768621886e-06,
+      "loss": 0.1585,
+      "step": 54900
+    },
+    {
+      "epoch": 0.7227456966889508,
+      "grad_norm": 2.0130560398101807,
+      "learning_rate": 1.8839793414483587e-06,
+      "loss": 0.1625,
+      "step": 54920
+    },
+    {
+      "epoch": 0.723008896141496,
+      "grad_norm": 1.7686140537261963,
+      "learning_rate": 1.8806471703179379e-06,
+      "loss": 0.1718,
+      "step": 54940
+    },
+    {
+      "epoch": 0.7232720955940412,
+      "grad_norm": 1.4748873710632324,
+      "learning_rate": 1.8773172658922773e-06,
+      "loss": 0.1783,
+      "step": 54960
+    },
+    {
+      "epoch": 0.7235352950465863,
+      "grad_norm": 2.013916015625,
+      "learning_rate": 1.8739896305910798e-06,
+      "loss": 0.1702,
+      "step": 54980
+    },
+    {
+      "epoch": 0.7237984944991315,
+      "grad_norm": 1.1692111492156982,
+      "learning_rate": 1.8706642668324e-06,
+      "loss": 0.1868,
+      "step": 55000
+    },
+    {
+      "epoch": 0.7237984944991315,
+      "eval_loss": 0.20438265800476074,
+      "eval_runtime": 3.8089,
+      "eval_samples_per_second": 4.201,
+      "eval_steps_per_second": 2.1,
+      "step": 55000
+    },
+    {
+      "epoch": 0.7240616939516766,
+      "grad_norm": 1.5731863975524902,
+      "learning_rate": 1.8673411770326466e-06,
+      "loss": 0.1699,
+      "step": 55020
+    },
+    {
+      "epoch": 0.7243248934042217,
+      "grad_norm": 1.2060843706130981,
+      "learning_rate": 1.864020363606565e-06,
+      "loss": 0.1809,
+      "step": 55040
+    },
+    {
+      "epoch": 0.7245880928567668,
+      "grad_norm": 1.7421520948410034,
+      "learning_rate": 1.8607018289672586e-06,
+      "loss": 0.1704,
+      "step": 55060
+    },
+    {
+      "epoch": 0.724851292309312,
+      "grad_norm": 1.4021812677383423,
+      "learning_rate": 1.8573855755261666e-06,
+      "loss": 0.1745,
+      "step": 55080
+    },
+    {
+      "epoch": 0.7251144917618572,
+      "grad_norm": 1.7057709693908691,
+      "learning_rate": 1.8540716056930747e-06,
+      "loss": 0.179,
+      "step": 55100
+    },
+    {
+      "epoch": 0.7253776912144023,
+      "grad_norm": 2.00797438621521,
+      "learning_rate": 1.8507599218761064e-06,
+      "loss": 0.1534,
+      "step": 55120
+    },
+    {
+      "epoch": 0.7256408906669474,
+      "grad_norm": 2.9509923458099365,
+      "learning_rate": 1.8474505264817244e-06,
+      "loss": 0.1587,
+      "step": 55140
+    },
+    {
+      "epoch": 0.7259040901194925,
+      "grad_norm": 2.007972478866577,
+      "learning_rate": 1.844143421914733e-06,
+      "loss": 0.177,
+      "step": 55160
+    },
+    {
+      "epoch": 0.7261672895720377,
+      "grad_norm": 2.121384620666504,
+      "learning_rate": 1.8408386105782689e-06,
+      "loss": 0.1707,
+      "step": 55180
+    },
+    {
+      "epoch": 0.7264304890245828,
+      "grad_norm": 1.665775179862976,
+      "learning_rate": 1.8375360948737952e-06,
+      "loss": 0.1756,
+      "step": 55200
+    },
+    {
+      "epoch": 0.726693688477128,
+      "grad_norm": 1.235682725906372,
+      "learning_rate": 1.8342358772011203e-06,
+      "loss": 0.1752,
+      "step": 55220
+    },
+    {
+      "epoch": 0.7269568879296731,
+      "grad_norm": 1.6774005889892578,
+      "learning_rate": 1.8309379599583732e-06,
+      "loss": 0.164,
+      "step": 55240
+    },
+    {
+      "epoch": 0.7272200873822182,
+      "grad_norm": 1.6855639219284058,
+      "learning_rate": 1.827642345542015e-06,
+      "loss": 0.1848,
+      "step": 55260
+    },
+    {
+      "epoch": 0.7274832868347634,
+      "grad_norm": 1.789559006690979,
+      "learning_rate": 1.8243490363468315e-06,
+      "loss": 0.1766,
+      "step": 55280
+    },
+    {
+      "epoch": 0.7277464862873085,
+      "grad_norm": 1.7778546810150146,
+      "learning_rate": 1.821058034765933e-06,
+      "loss": 0.1761,
+      "step": 55300
+    },
+    {
+      "epoch": 0.7280096857398537,
+      "grad_norm": 2.164508819580078,
+      "learning_rate": 1.8177693431907578e-06,
+      "loss": 0.1757,
+      "step": 55320
+    },
+    {
+      "epoch": 0.7282728851923987,
+      "grad_norm": 1.5303072929382324,
+      "learning_rate": 1.8144829640110601e-06,
+      "loss": 0.1711,
+      "step": 55340
+    },
+    {
+      "epoch": 0.7285360846449439,
+      "grad_norm": 1.4970812797546387,
+      "learning_rate": 1.8111988996149165e-06,
+      "loss": 0.1634,
+      "step": 55360
+    },
+    {
+      "epoch": 0.7287992840974891,
+      "grad_norm": 1.2830289602279663,
+      "learning_rate": 1.8079171523887206e-06,
+      "loss": 0.1684,
+      "step": 55380
+    },
+    {
+      "epoch": 0.7290624835500342,
+      "grad_norm": 1.8833084106445312,
+      "learning_rate": 1.8046377247171804e-06,
+      "loss": 0.174,
+      "step": 55400
+    },
+    {
+      "epoch": 0.7293256830025794,
+      "grad_norm": 1.760806679725647,
+      "learning_rate": 1.8013606189833255e-06,
+      "loss": 0.1761,
+      "step": 55420
+    },
+    {
+      "epoch": 0.7295888824551245,
+      "grad_norm": 1.6063445806503296,
+      "learning_rate": 1.7980858375684885e-06,
+      "loss": 0.1755,
+      "step": 55440
+    },
+    {
+      "epoch": 0.7298520819076696,
+      "grad_norm": 1.5186870098114014,
+      "learning_rate": 1.7948133828523178e-06,
+      "loss": 0.1786,
+      "step": 55460
+    },
+    {
+      "epoch": 0.7301152813602148,
+      "grad_norm": 1.7916417121887207,
+      "learning_rate": 1.7915432572127739e-06,
+      "loss": 0.1722,
+      "step": 55480
+    },
+    {
+      "epoch": 0.7303784808127599,
+      "grad_norm": 1.7719175815582275,
+      "learning_rate": 1.78827546302612e-06,
+      "loss": 0.1711,
+      "step": 55500
+    },
+    {
+      "epoch": 0.7303784808127599,
+      "eval_loss": 0.2098965346813202,
+      "eval_runtime": 4.3445,
+      "eval_samples_per_second": 3.683,
+      "eval_steps_per_second": 1.841,
+      "step": 55500
+    },
+    {
+      "epoch": 0.7306416802653051,
+      "grad_norm": 1.2200742959976196,
+      "learning_rate": 1.7850100026669276e-06,
+      "loss": 0.155,
+      "step": 55520
+    },
+    {
+      "epoch": 0.7309048797178502,
+      "grad_norm": 1.2758326530456543,
+      "learning_rate": 1.7817468785080722e-06,
+      "loss": 0.194,
+      "step": 55540
+    },
+    {
+      "epoch": 0.7311680791703953,
+      "grad_norm": 2.2209832668304443,
+      "learning_rate": 1.7784860929207281e-06,
+      "loss": 0.1778,
+      "step": 55560
+    },
+    {
+      "epoch": 0.7314312786229404,
+      "grad_norm": 1.3274725675582886,
+      "learning_rate": 1.7752276482743786e-06,
+      "loss": 0.1808,
+      "step": 55580
+    },
+    {
+      "epoch": 0.7316944780754856,
+      "grad_norm": 1.6828707456588745,
+      "learning_rate": 1.7719715469367992e-06,
+      "loss": 0.1606,
+      "step": 55600
+    },
+    {
+      "epoch": 0.7319576775280308,
+      "grad_norm": 2.0809922218322754,
+      "learning_rate": 1.7687177912740594e-06,
+      "loss": 0.1624,
+      "step": 55620
+    },
+    {
+      "epoch": 0.7322208769805759,
+      "grad_norm": 1.410866141319275,
+      "learning_rate": 1.7654663836505349e-06,
+      "loss": 0.1709,
+      "step": 55640
+    },
+    {
+      "epoch": 0.732484076433121,
+      "grad_norm": 1.4719336032867432,
+      "learning_rate": 1.762217326428885e-06,
+      "loss": 0.1744,
+      "step": 55660
+    },
+    {
+      "epoch": 0.7327472758856661,
+      "grad_norm": 1.3056925535202026,
+      "learning_rate": 1.7589706219700703e-06,
+      "loss": 0.1793,
+      "step": 55680
+    },
+    {
+      "epoch": 0.7330104753382113,
+      "grad_norm": 1.7468953132629395,
+      "learning_rate": 1.7557262726333318e-06,
+      "loss": 0.1701,
+      "step": 55700
+    },
+    {
+      "epoch": 0.7332736747907564,
+      "grad_norm": 1.2766671180725098,
+      "learning_rate": 1.752484280776205e-06,
+      "loss": 0.1671,
+      "step": 55720
+    },
+    {
+      "epoch": 0.7335368742433016,
+      "grad_norm": 1.4504164457321167,
+      "learning_rate": 1.7492446487545133e-06,
+      "loss": 0.1712,
+      "step": 55740
+    },
+    {
+      "epoch": 0.7338000736958468,
+      "grad_norm": 1.81706964969635,
+      "learning_rate": 1.7460073789223625e-06,
+      "loss": 0.1667,
+      "step": 55760
+    },
+    {
+      "epoch": 0.7340632731483918,
+      "grad_norm": 1.7889294624328613,
+      "learning_rate": 1.7427724736321422e-06,
+      "loss": 0.1755,
+      "step": 55780
+    },
+    {
+      "epoch": 0.734326472600937,
+      "grad_norm": 1.4291560649871826,
+      "learning_rate": 1.739539935234525e-06,
+      "loss": 0.1904,
+      "step": 55800
+    },
+    {
+      "epoch": 0.7345896720534821,
+      "grad_norm": 1.8920379877090454,
+      "learning_rate": 1.7363097660784612e-06,
+      "loss": 0.1688,
+      "step": 55820
+    },
+    {
+      "epoch": 0.7348528715060273,
+      "grad_norm": 1.6893349885940552,
+      "learning_rate": 1.733081968511186e-06,
+      "loss": 0.1716,
+      "step": 55840
+    },
+    {
+      "epoch": 0.7351160709585725,
+      "grad_norm": 2.1283702850341797,
+      "learning_rate": 1.7298565448782013e-06,
+      "loss": 0.1984,
+      "step": 55860
+    },
+    {
+      "epoch": 0.7353792704111175,
+      "grad_norm": 1.4097987413406372,
+      "learning_rate": 1.726633497523289e-06,
+      "loss": 0.1749,
+      "step": 55880
+    },
+    {
+      "epoch": 0.7356424698636627,
+      "grad_norm": 2.0915513038635254,
+      "learning_rate": 1.7234128287885087e-06,
+      "loss": 0.1756,
+      "step": 55900
+    },
+    {
+      "epoch": 0.7359056693162078,
+      "grad_norm": 1.147087574005127,
+      "learning_rate": 1.7201945410141847e-06,
+      "loss": 0.1784,
+      "step": 55920
+    },
+    {
+      "epoch": 0.736168868768753,
+      "grad_norm": 1.6151602268218994,
+      "learning_rate": 1.7169786365389135e-06,
+      "loss": 0.1839,
+      "step": 55940
+    },
+    {
+      "epoch": 0.7364320682212981,
+      "grad_norm": 1.4529964923858643,
+      "learning_rate": 1.7137651176995607e-06,
+      "loss": 0.1685,
+      "step": 55960
+    },
+    {
+      "epoch": 0.7366952676738432,
+      "grad_norm": 1.3683630228042603,
+      "learning_rate": 1.7105539868312554e-06,
+      "loss": 0.1637,
+      "step": 55980
+    },
+    {
+      "epoch": 0.7369584671263884,
+      "grad_norm": 1.2874741554260254,
+      "learning_rate": 1.7073452462673972e-06,
+      "loss": 0.1729,
+      "step": 56000
+    },
+    {
+      "epoch": 0.7369584671263884,
+      "eval_loss": 0.1988041251897812,
+      "eval_runtime": 4.4105,
+      "eval_samples_per_second": 3.628,
+      "eval_steps_per_second": 1.814,
+      "step": 56000
+    },
+    {
+      "epoch": 0.7372216665789335,
+      "grad_norm": 1.6109411716461182,
+      "learning_rate": 1.7041388983396456e-06,
+      "loss": 0.1865,
+      "step": 56020
+    },
+    {
+      "epoch": 0.7374848660314787,
+      "grad_norm": 1.7081708908081055,
+      "learning_rate": 1.700934945377915e-06,
+      "loss": 0.1594,
+      "step": 56040
+    },
+    {
+      "epoch": 0.7377480654840238,
+      "grad_norm": 1.2817937135696411,
+      "learning_rate": 1.697733389710392e-06,
+      "loss": 0.1738,
+      "step": 56060
+    },
+    {
+      "epoch": 0.7380112649365689,
+      "grad_norm": 1.406073808670044,
+      "learning_rate": 1.6945342336635118e-06,
+      "loss": 0.1578,
+      "step": 56080
+    },
+    {
+      "epoch": 0.738274464389114,
+      "grad_norm": 1.4198691844940186,
+      "learning_rate": 1.6913374795619696e-06,
+      "loss": 0.1691,
+      "step": 56100
+    },
+    {
+      "epoch": 0.7385376638416592,
+      "grad_norm": 1.740362524986267,
+      "learning_rate": 1.6881431297287148e-06,
+      "loss": 0.1809,
+      "step": 56120
+    },
+    {
+      "epoch": 0.7388008632942044,
+      "grad_norm": 1.8833866119384766,
+      "learning_rate": 1.6849511864849466e-06,
+      "loss": 0.1838,
+      "step": 56140
+    },
+    {
+      "epoch": 0.7390640627467495,
+      "grad_norm": 1.8501611948013306,
+      "learning_rate": 1.6817616521501228e-06,
+      "loss": 0.1811,
+      "step": 56160
+    },
+    {
+      "epoch": 0.7393272621992947,
+      "grad_norm": 1.6484451293945312,
+      "learning_rate": 1.6785745290419435e-06,
+      "loss": 0.1541,
+      "step": 56180
+    },
+    {
+      "epoch": 0.7395904616518397,
+      "grad_norm": 1.6441936492919922,
+      "learning_rate": 1.6753898194763596e-06,
+      "loss": 0.1866,
+      "step": 56200
+    },
+    {
+      "epoch": 0.7398536611043849,
+      "grad_norm": 1.3877925872802734,
+      "learning_rate": 1.6722075257675678e-06,
+      "loss": 0.2022,
+      "step": 56220
+    },
+    {
+      "epoch": 0.74011686055693,
+      "grad_norm": 1.3088642358779907,
+      "learning_rate": 1.6690276502280083e-06,
+      "loss": 0.159,
+      "step": 56240
+    },
+    {
+      "epoch": 0.7403800600094752,
+      "grad_norm": 1.2424051761627197,
+      "learning_rate": 1.6658501951683703e-06,
+      "loss": 0.1843,
+      "step": 56260
+    },
+    {
+      "epoch": 0.7406432594620204,
+      "grad_norm": 1.6790339946746826,
+      "learning_rate": 1.662675162897573e-06,
+      "loss": 0.1651,
+      "step": 56280
+    },
+    {
+      "epoch": 0.7409064589145654,
+      "grad_norm": 2.1148552894592285,
+      "learning_rate": 1.6595025557227817e-06,
+      "loss": 0.1642,
+      "step": 56300
+    },
+    {
+      "epoch": 0.7411696583671106,
+      "grad_norm": 1.140949010848999,
+      "learning_rate": 1.6563323759494021e-06,
+      "loss": 0.1818,
+      "step": 56320
+    },
+    {
+      "epoch": 0.7414328578196557,
+      "grad_norm": 1.501089334487915,
+      "learning_rate": 1.6531646258810714e-06,
+      "loss": 0.1675,
+      "step": 56340
+    },
+    {
+      "epoch": 0.7416960572722009,
+      "grad_norm": 1.543915867805481,
+      "learning_rate": 1.6499993078196624e-06,
+      "loss": 0.1602,
+      "step": 56360
+    },
+    {
+      "epoch": 0.741959256724746,
+      "grad_norm": 1.6834274530410767,
+      "learning_rate": 1.6468364240652813e-06,
+      "loss": 0.1729,
+      "step": 56380
+    },
+    {
+      "epoch": 0.7422224561772911,
+      "grad_norm": 1.7343764305114746,
+      "learning_rate": 1.6436759769162626e-06,
+      "loss": 0.167,
+      "step": 56400
+    },
+    {
+      "epoch": 0.7424856556298363,
+      "grad_norm": 1.2986983060836792,
+      "learning_rate": 1.640517968669178e-06,
+      "loss": 0.17,
+      "step": 56420
+    },
+    {
+      "epoch": 0.7427488550823814,
+      "grad_norm": 1.537785530090332,
+      "learning_rate": 1.637362401618815e-06,
+      "loss": 0.1514,
+      "step": 56440
+    },
+    {
+      "epoch": 0.7430120545349266,
+      "grad_norm": 2.075199842453003,
+      "learning_rate": 1.6342092780581985e-06,
+      "loss": 0.1592,
+      "step": 56460
+    },
+    {
+      "epoch": 0.7432752539874717,
+      "grad_norm": 1.9427604675292969,
+      "learning_rate": 1.6310586002785717e-06,
+      "loss": 0.1586,
+      "step": 56480
+    },
+    {
+      "epoch": 0.7435384534400169,
+      "grad_norm": 2.031938076019287,
+      "learning_rate": 1.6279103705694016e-06,
+      "loss": 0.1727,
+      "step": 56500
+    },
+    {
+      "epoch": 0.7435384534400169,
+      "eval_loss": 0.20653128623962402,
+      "eval_runtime": 4.3419,
+      "eval_samples_per_second": 3.685,
+      "eval_steps_per_second": 1.842,
+      "step": 56500
+    },
+    {
+      "epoch": 0.743801652892562,
+      "grad_norm": 2.1238136291503906,
+      "learning_rate": 1.6247645912183768e-06,
+      "loss": 0.17,
+      "step": 56520
+    },
+    {
+      "epoch": 0.7440648523451071,
+      "grad_norm": 1.6529401540756226,
+      "learning_rate": 1.6216212645114033e-06,
+      "loss": 0.1659,
+      "step": 56540
+    },
+    {
+      "epoch": 0.7443280517976523,
+      "grad_norm": 1.4737370014190674,
+      "learning_rate": 1.6184803927326099e-06,
+      "loss": 0.1773,
+      "step": 56560
+    },
+    {
+      "epoch": 0.7445912512501974,
+      "grad_norm": 1.252873182296753,
+      "learning_rate": 1.615341978164337e-06,
+      "loss": 0.1812,
+      "step": 56580
+    },
+    {
+      "epoch": 0.7448544507027426,
+      "grad_norm": 1.411291480064392,
+      "learning_rate": 1.6122060230871406e-06,
+      "loss": 0.169,
+      "step": 56600
+    },
+    {
+      "epoch": 0.7451176501552876,
+      "grad_norm": 1.706101417541504,
+      "learning_rate": 1.6090725297797893e-06,
+      "loss": 0.1777,
+      "step": 56620
+    },
+    {
+      "epoch": 0.7453808496078328,
+      "grad_norm": 1.7191081047058105,
+      "learning_rate": 1.6059415005192635e-06,
+      "loss": 0.1673,
+      "step": 56640
+    },
+    {
+      "epoch": 0.745644049060378,
+      "grad_norm": 1.508251667022705,
+      "learning_rate": 1.6028129375807512e-06,
+      "loss": 0.1586,
+      "step": 56660
+    },
+    {
+      "epoch": 0.7459072485129231,
+      "grad_norm": 1.7575926780700684,
+      "learning_rate": 1.5996868432376545e-06,
+      "loss": 0.1701,
+      "step": 56680
+    },
+    {
+      "epoch": 0.7461704479654683,
+      "grad_norm": 1.7897168397903442,
+      "learning_rate": 1.5965632197615705e-06,
+      "loss": 0.166,
+      "step": 56700
+    },
+    {
+      "epoch": 0.7464336474180133,
+      "grad_norm": 1.7012605667114258,
+      "learning_rate": 1.5934420694223119e-06,
+      "loss": 0.1758,
+      "step": 56720
+    },
+    {
+      "epoch": 0.7466968468705585,
+      "grad_norm": 2.173084020614624,
+      "learning_rate": 1.5903233944878876e-06,
+      "loss": 0.1745,
+      "step": 56740
+    },
+    {
+      "epoch": 0.7469600463231036,
+      "grad_norm": 1.668166160583496,
+      "learning_rate": 1.5872071972245111e-06,
+      "loss": 0.1562,
+      "step": 56760
+    },
+    {
+      "epoch": 0.7472232457756488,
+      "grad_norm": 1.6727999448776245,
+      "learning_rate": 1.5840934798965928e-06,
+      "loss": 0.1531,
+      "step": 56780
+    },
+    {
+      "epoch": 0.747486445228194,
+      "grad_norm": 1.698733925819397,
+      "learning_rate": 1.5809822447667422e-06,
+      "loss": 0.1714,
+      "step": 56800
+    },
+    {
+      "epoch": 0.747749644680739,
+      "grad_norm": 1.4367616176605225,
+      "learning_rate": 1.5778734940957668e-06,
+      "loss": 0.1753,
+      "step": 56820
+    },
+    {
+      "epoch": 0.7480128441332842,
+      "grad_norm": 1.700091004371643,
+      "learning_rate": 1.5747672301426682e-06,
+      "loss": 0.1768,
+      "step": 56840
+    },
+    {
+      "epoch": 0.7482760435858293,
+      "grad_norm": 1.4870096445083618,
+      "learning_rate": 1.5716634551646353e-06,
+      "loss": 0.1614,
+      "step": 56860
+    },
+    {
+      "epoch": 0.7485392430383745,
+      "grad_norm": 1.4967890977859497,
+      "learning_rate": 1.5685621714170574e-06,
+      "loss": 0.1756,
+      "step": 56880
+    },
+    {
+      "epoch": 0.7488024424909197,
+      "grad_norm": 1.7811858654022217,
+      "learning_rate": 1.5654633811535075e-06,
+      "loss": 0.1573,
+      "step": 56900
+    },
+    {
+      "epoch": 0.7490656419434648,
+      "grad_norm": 1.6797267198562622,
+      "learning_rate": 1.5623670866257495e-06,
+      "loss": 0.1757,
+      "step": 56920
+    },
+    {
+      "epoch": 0.7493288413960099,
+      "grad_norm": 1.7019544839859009,
+      "learning_rate": 1.559273290083731e-06,
+      "loss": 0.1819,
+      "step": 56940
+    },
+    {
+      "epoch": 0.749592040848555,
+      "grad_norm": 1.7676286697387695,
+      "learning_rate": 1.5561819937755858e-06,
+      "loss": 0.1923,
+      "step": 56960
+    },
+    {
+      "epoch": 0.7498552403011002,
+      "grad_norm": 1.7360838651657104,
+      "learning_rate": 1.553093199947634e-06,
+      "loss": 0.1734,
+      "step": 56980
+    },
+    {
+      "epoch": 0.7501184397536453,
+      "grad_norm": 2.004392147064209,
+      "learning_rate": 1.550006910844372e-06,
+      "loss": 0.1684,
+      "step": 57000
+    },
+    {
+      "epoch": 0.7501184397536453,
+      "eval_loss": 0.20283523201942444,
+      "eval_runtime": 4.4315,
+      "eval_samples_per_second": 3.611,
+      "eval_steps_per_second": 1.805,
+      "step": 57000
+    },
+    {
+      "epoch": 0.7503816392061905,
+      "grad_norm": 1.875972867012024,
+      "learning_rate": 1.5469231287084795e-06,
+      "loss": 0.1687,
+      "step": 57020
+    },
+    {
+      "epoch": 0.7506448386587355,
+      "grad_norm": 1.9650957584381104,
+      "learning_rate": 1.5438418557808139e-06,
+      "loss": 0.173,
+      "step": 57040
+    },
+    {
+      "epoch": 0.7509080381112807,
+      "grad_norm": 1.0947922468185425,
+      "learning_rate": 1.5407630943004054e-06,
+      "loss": 0.1705,
+      "step": 57060
+    },
+    {
+      "epoch": 0.7511712375638259,
+      "grad_norm": 1.6334928274154663,
+      "learning_rate": 1.5376868465044692e-06,
+      "loss": 0.1605,
+      "step": 57080
+    },
+    {
+      "epoch": 0.751434437016371,
+      "grad_norm": 2.395334243774414,
+      "learning_rate": 1.5346131146283822e-06,
+      "loss": 0.171,
+      "step": 57100
+    },
+    {
+      "epoch": 0.7516976364689162,
+      "grad_norm": 1.6541242599487305,
+      "learning_rate": 1.531541900905698e-06,
+      "loss": 0.155,
+      "step": 57120
+    },
+    {
+      "epoch": 0.7519608359214612,
+      "grad_norm": 2.4164140224456787,
+      "learning_rate": 1.5284732075681441e-06,
+      "loss": 0.1719,
+      "step": 57140
+    },
+    {
+      "epoch": 0.7522240353740064,
+      "grad_norm": 1.2857273817062378,
+      "learning_rate": 1.5254070368456114e-06,
+      "loss": 0.1828,
+      "step": 57160
+    },
+    {
+      "epoch": 0.7524872348265516,
+      "grad_norm": 1.498409390449524,
+      "learning_rate": 1.5223433909661595e-06,
+      "loss": 0.1831,
+      "step": 57180
+    },
+    {
+      "epoch": 0.7527504342790967,
+      "grad_norm": 1.8883649110794067,
+      "learning_rate": 1.519282272156013e-06,
+      "loss": 0.1699,
+      "step": 57200
+    },
+    {
+      "epoch": 0.7530136337316419,
+      "grad_norm": 1.5593371391296387,
+      "learning_rate": 1.5162236826395587e-06,
+      "loss": 0.1726,
+      "step": 57220
+    },
+    {
+      "epoch": 0.7532768331841869,
+      "grad_norm": 1.4707415103912354,
+      "learning_rate": 1.5131676246393511e-06,
+      "loss": 0.1547,
+      "step": 57240
+    },
+    {
+      "epoch": 0.7535400326367321,
+      "grad_norm": 1.7885115146636963,
+      "learning_rate": 1.5101141003761005e-06,
+      "loss": 0.1693,
+      "step": 57260
+    },
+    {
+      "epoch": 0.7538032320892772,
+      "grad_norm": 1.5300804376602173,
+      "learning_rate": 1.5070631120686723e-06,
+      "loss": 0.1601,
+      "step": 57280
+    },
+    {
+      "epoch": 0.7540664315418224,
+      "grad_norm": 1.574967384338379,
+      "learning_rate": 1.5040146619340985e-06,
+      "loss": 0.1686,
+      "step": 57300
+    },
+    {
+      "epoch": 0.7543296309943676,
+      "grad_norm": 1.379263997077942,
+      "learning_rate": 1.5009687521875588e-06,
+      "loss": 0.1752,
+      "step": 57320
+    },
+    {
+      "epoch": 0.7545928304469127,
+      "grad_norm": 1.9187061786651611,
+      "learning_rate": 1.4979253850423935e-06,
+      "loss": 0.185,
+      "step": 57340
+    },
+    {
+      "epoch": 0.7548560298994578,
+      "grad_norm": 1.4053794145584106,
+      "learning_rate": 1.4948845627100882e-06,
+      "loss": 0.1784,
+      "step": 57360
+    },
+    {
+      "epoch": 0.7551192293520029,
+      "grad_norm": 1.6748844385147095,
+      "learning_rate": 1.491846287400282e-06,
+      "loss": 0.1693,
+      "step": 57380
+    },
+    {
+      "epoch": 0.7553824288045481,
+      "grad_norm": 1.9139264822006226,
+      "learning_rate": 1.4888105613207677e-06,
+      "loss": 0.1627,
+      "step": 57400
+    },
+    {
+      "epoch": 0.7556456282570932,
+      "grad_norm": 2.036015033721924,
+      "learning_rate": 1.4857773866774804e-06,
+      "loss": 0.1751,
+      "step": 57420
+    },
+    {
+      "epoch": 0.7559088277096384,
+      "grad_norm": 1.1797986030578613,
+      "learning_rate": 1.482746765674502e-06,
+      "loss": 0.1639,
+      "step": 57440
+    },
+    {
+      "epoch": 0.7561720271621835,
+      "grad_norm": 1.7978408336639404,
+      "learning_rate": 1.4797187005140606e-06,
+      "loss": 0.1784,
+      "step": 57460
+    },
+    {
+      "epoch": 0.7564352266147286,
+      "grad_norm": 1.5723628997802734,
+      "learning_rate": 1.4766931933965234e-06,
+      "loss": 0.1688,
+      "step": 57480
+    },
+    {
+      "epoch": 0.7566984260672738,
+      "grad_norm": 1.3241543769836426,
+      "learning_rate": 1.4736702465204067e-06,
+      "loss": 0.159,
+      "step": 57500
+    },
+    {
+      "epoch": 0.7566984260672738,
+      "eval_loss": 0.20226024091243744,
+      "eval_runtime": 4.3963,
+      "eval_samples_per_second": 3.639,
+      "eval_steps_per_second": 1.82,
+      "step": 57500
+    },
+    {
+      "epoch": 0.7569616255198189,
+      "grad_norm": 1.4160575866699219,
+      "learning_rate": 1.4706498620823561e-06,
+      "loss": 0.1851,
+      "step": 57520
+    },
+    {
+      "epoch": 0.7572248249723641,
+      "grad_norm": 1.9417935609817505,
+      "learning_rate": 1.46763204227716e-06,
+      "loss": 0.1757,
+      "step": 57540
+    },
+    {
+      "epoch": 0.7574880244249091,
+      "grad_norm": 1.4732075929641724,
+      "learning_rate": 1.4646167892977464e-06,
+      "loss": 0.1814,
+      "step": 57560
+    },
+    {
+      "epoch": 0.7577512238774543,
+      "grad_norm": 1.4541707038879395,
+      "learning_rate": 1.4616041053351738e-06,
+      "loss": 0.1886,
+      "step": 57580
+    },
+    {
+      "epoch": 0.7580144233299995,
+      "grad_norm": 1.2551279067993164,
+      "learning_rate": 1.4585939925786352e-06,
+      "loss": 0.1677,
+      "step": 57600
+    },
+    {
+      "epoch": 0.7582776227825446,
+      "grad_norm": 1.687118649482727,
+      "learning_rate": 1.4555864532154544e-06,
+      "loss": 0.1853,
+      "step": 57620
+    },
+    {
+      "epoch": 0.7585408222350898,
+      "grad_norm": 1.8394262790679932,
+      "learning_rate": 1.4525814894310853e-06,
+      "loss": 0.1772,
+      "step": 57640
+    },
+    {
+      "epoch": 0.7588040216876349,
+      "grad_norm": 1.2318722009658813,
+      "learning_rate": 1.4495791034091129e-06,
+      "loss": 0.1643,
+      "step": 57660
+    },
+    {
+      "epoch": 0.75906722114018,
+      "grad_norm": 1.7091741561889648,
+      "learning_rate": 1.4465792973312481e-06,
+      "loss": 0.1775,
+      "step": 57680
+    },
+    {
+      "epoch": 0.7593304205927252,
+      "grad_norm": 1.5151634216308594,
+      "learning_rate": 1.443582073377321e-06,
+      "loss": 0.171,
+      "step": 57700
+    },
+    {
+      "epoch": 0.7595936200452703,
+      "grad_norm": 1.7271569967269897,
+      "learning_rate": 1.440587433725294e-06,
+      "loss": 0.1663,
+      "step": 57720
+    },
+    {
+      "epoch": 0.7598568194978155,
+      "grad_norm": 1.781137466430664,
+      "learning_rate": 1.4375953805512478e-06,
+      "loss": 0.1656,
+      "step": 57740
+    },
+    {
+      "epoch": 0.7601200189503606,
+      "grad_norm": 1.8237067461013794,
+      "learning_rate": 1.4346059160293834e-06,
+      "loss": 0.1682,
+      "step": 57760
+    },
+    {
+      "epoch": 0.7603832184029057,
+      "grad_norm": 1.4599294662475586,
+      "learning_rate": 1.4316190423320202e-06,
+      "loss": 0.1625,
+      "step": 57780
+    },
+    {
+      "epoch": 0.7606464178554508,
+      "grad_norm": 1.8882577419281006,
+      "learning_rate": 1.428634761629596e-06,
+      "loss": 0.1849,
+      "step": 57800
+    },
+    {
+      "epoch": 0.760909617307996,
+      "grad_norm": 1.9581879377365112,
+      "learning_rate": 1.4256530760906656e-06,
+      "loss": 0.1773,
+      "step": 57820
+    },
+    {
+      "epoch": 0.7611728167605412,
+      "grad_norm": 1.614025592803955,
+      "learning_rate": 1.4226739878818957e-06,
+      "loss": 0.1596,
+      "step": 57840
+    },
+    {
+      "epoch": 0.7614360162130863,
+      "grad_norm": 1.6892732381820679,
+      "learning_rate": 1.4196974991680667e-06,
+      "loss": 0.1614,
+      "step": 57860
+    },
+    {
+      "epoch": 0.7616992156656314,
+      "grad_norm": 1.9163165092468262,
+      "learning_rate": 1.4167236121120703e-06,
+      "loss": 0.1713,
+      "step": 57880
+    },
+    {
+      "epoch": 0.7619624151181765,
+      "grad_norm": 1.5661357641220093,
+      "learning_rate": 1.413752328874905e-06,
+      "loss": 0.1603,
+      "step": 57900
+    },
+    {
+      "epoch": 0.7622256145707217,
+      "grad_norm": 1.7975029945373535,
+      "learning_rate": 1.4107836516156847e-06,
+      "loss": 0.181,
+      "step": 57920
+    },
+    {
+      "epoch": 0.7624888140232668,
+      "grad_norm": 1.5743542909622192,
+      "learning_rate": 1.4078175824916178e-06,
+      "loss": 0.1943,
+      "step": 57940
+    },
+    {
+      "epoch": 0.762752013475812,
+      "grad_norm": 1.7113069295883179,
+      "learning_rate": 1.4048541236580294e-06,
+      "loss": 0.1735,
+      "step": 57960
+    },
+    {
+      "epoch": 0.7630152129283571,
+      "grad_norm": 1.4355137348175049,
+      "learning_rate": 1.4018932772683402e-06,
+      "loss": 0.1953,
+      "step": 57980
+    },
+    {
+      "epoch": 0.7632784123809022,
+      "grad_norm": 1.2221993207931519,
+      "learning_rate": 1.3989350454740753e-06,
+      "loss": 0.1579,
+      "step": 58000
+    },
+    {
+      "epoch": 0.7632784123809022,
+      "eval_loss": 0.20907804369926453,
+      "eval_runtime": 3.3809,
+      "eval_samples_per_second": 4.732,
+      "eval_steps_per_second": 2.366,
+      "step": 58000
+    },
+    {
+      "epoch": 0.7635416118334474,
+      "grad_norm": 1.1575959920883179,
+      "learning_rate": 1.3959794304248596e-06,
+      "loss": 0.1609,
+      "step": 58020
+    },
+    {
+      "epoch": 0.7638048112859925,
+      "grad_norm": 2.0537211894989014,
+      "learning_rate": 1.3930264342684163e-06,
+      "loss": 0.1637,
+      "step": 58040
+    },
+    {
+      "epoch": 0.7640680107385377,
+      "grad_norm": 1.7187657356262207,
+      "learning_rate": 1.3900760591505636e-06,
+      "loss": 0.1638,
+      "step": 58060
+    },
+    {
+      "epoch": 0.7643312101910829,
+      "grad_norm": 1.4475919008255005,
+      "learning_rate": 1.3871283072152225e-06,
+      "loss": 0.1704,
+      "step": 58080
+    },
+    {
+      "epoch": 0.7645944096436279,
+      "grad_norm": 1.7150938510894775,
+      "learning_rate": 1.384183180604396e-06,
+      "loss": 0.1723,
+      "step": 58100
+    },
+    {
+      "epoch": 0.7648576090961731,
+      "grad_norm": 1.4818812608718872,
+      "learning_rate": 1.3812406814581903e-06,
+      "loss": 0.1647,
+      "step": 58120
+    },
+    {
+      "epoch": 0.7651208085487182,
+      "grad_norm": 1.9546701908111572,
+      "learning_rate": 1.3783008119147967e-06,
+      "loss": 0.1797,
+      "step": 58140
+    },
+    {
+      "epoch": 0.7653840080012634,
+      "grad_norm": 1.438045859336853,
+      "learning_rate": 1.3753635741104971e-06,
+      "loss": 0.1956,
+      "step": 58160
+    },
+    {
+      "epoch": 0.7656472074538085,
+      "grad_norm": 1.4760079383850098,
+      "learning_rate": 1.37242897017966e-06,
+      "loss": 0.1653,
+      "step": 58180
+    },
+    {
+      "epoch": 0.7659104069063536,
+      "grad_norm": 1.2521904706954956,
+      "learning_rate": 1.36949700225474e-06,
+      "loss": 0.1695,
+      "step": 58200
+    },
+    {
+      "epoch": 0.7661736063588988,
+      "grad_norm": 1.7402939796447754,
+      "learning_rate": 1.3665676724662792e-06,
+      "loss": 0.1662,
+      "step": 58220
+    },
+    {
+      "epoch": 0.7664368058114439,
+      "grad_norm": 1.7938235998153687,
+      "learning_rate": 1.3636409829428999e-06,
+      "loss": 0.1718,
+      "step": 58240
+    },
+    {
+      "epoch": 0.7667000052639891,
+      "grad_norm": 1.5610918998718262,
+      "learning_rate": 1.3607169358113054e-06,
+      "loss": 0.1762,
+      "step": 58260
+    },
+    {
+      "epoch": 0.7669632047165342,
+      "grad_norm": 1.6662626266479492,
+      "learning_rate": 1.3577955331962812e-06,
+      "loss": 0.1938,
+      "step": 58280
+    },
+    {
+      "epoch": 0.7672264041690793,
+      "grad_norm": 1.4707192182540894,
+      "learning_rate": 1.3548767772206884e-06,
+      "loss": 0.184,
+      "step": 58300
+    },
+    {
+      "epoch": 0.7674896036216244,
+      "grad_norm": 1.9221549034118652,
+      "learning_rate": 1.351960670005466e-06,
+      "loss": 0.1751,
+      "step": 58320
+    },
+    {
+      "epoch": 0.7677528030741696,
+      "grad_norm": 1.8058050870895386,
+      "learning_rate": 1.3490472136696325e-06,
+      "loss": 0.1592,
+      "step": 58340
+    },
+    {
+      "epoch": 0.7680160025267148,
+      "grad_norm": 2.015887498855591,
+      "learning_rate": 1.3461364103302699e-06,
+      "loss": 0.1904,
+      "step": 58360
+    },
+    {
+      "epoch": 0.7682792019792599,
+      "grad_norm": 1.6276799440383911,
+      "learning_rate": 1.3432282621025439e-06,
+      "loss": 0.1619,
+      "step": 58380
+    },
+    {
+      "epoch": 0.768542401431805,
+      "grad_norm": 1.2927327156066895,
+      "learning_rate": 1.3403227710996835e-06,
+      "loss": 0.1699,
+      "step": 58400
+    },
+    {
+      "epoch": 0.7688056008843501,
+      "grad_norm": 1.3424079418182373,
+      "learning_rate": 1.337419939432989e-06,
+      "loss": 0.1652,
+      "step": 58420
+    },
+    {
+      "epoch": 0.7690688003368953,
+      "grad_norm": 1.899392008781433,
+      "learning_rate": 1.3345197692118285e-06,
+      "loss": 0.1638,
+      "step": 58440
+    },
+    {
+      "epoch": 0.7693319997894404,
+      "grad_norm": 1.728190302848816,
+      "learning_rate": 1.3316222625436342e-06,
+      "loss": 0.1812,
+      "step": 58460
+    },
+    {
+      "epoch": 0.7695951992419856,
+      "grad_norm": 1.395731806755066,
+      "learning_rate": 1.3287274215339074e-06,
+      "loss": 0.1565,
+      "step": 58480
+    },
+    {
+      "epoch": 0.7698583986945308,
+      "grad_norm": 1.7818132638931274,
+      "learning_rate": 1.325835248286209e-06,
+      "loss": 0.1823,
+      "step": 58500
+    },
+    {
+      "epoch": 0.7698583986945308,
+      "eval_loss": 0.20271533727645874,
+      "eval_runtime": 4.5061,
+      "eval_samples_per_second": 3.551,
+      "eval_steps_per_second": 1.775,
+      "step": 58500
+    },
+    {
+      "epoch": 0.7701215981470758,
+      "grad_norm": 1.7293717861175537,
+      "learning_rate": 1.3229457449021583e-06,
+      "loss": 0.1862,
+      "step": 58520
+    },
+    {
+      "epoch": 0.770384797599621,
+      "grad_norm": 1.7409193515777588,
+      "learning_rate": 1.3200589134814412e-06,
+      "loss": 0.1831,
+      "step": 58540
+    },
+    {
+      "epoch": 0.7706479970521661,
+      "grad_norm": 2.041299343109131,
+      "learning_rate": 1.3171747561217978e-06,
+      "loss": 0.1773,
+      "step": 58560
+    },
+    {
+      "epoch": 0.7709111965047113,
+      "grad_norm": 1.7543590068817139,
+      "learning_rate": 1.3142932749190263e-06,
+      "loss": 0.1639,
+      "step": 58580
+    },
+    {
+      "epoch": 0.7711743959572565,
+      "grad_norm": 1.7173969745635986,
+      "learning_rate": 1.3114144719669797e-06,
+      "loss": 0.1701,
+      "step": 58600
+    },
+    {
+      "epoch": 0.7714375954098015,
+      "grad_norm": 1.8246970176696777,
+      "learning_rate": 1.3085383493575637e-06,
+      "loss": 0.1872,
+      "step": 58620
+    },
+    {
+      "epoch": 0.7717007948623467,
+      "grad_norm": 1.8658499717712402,
+      "learning_rate": 1.3056649091807404e-06,
+      "loss": 0.1574,
+      "step": 58640
+    },
+    {
+      "epoch": 0.7719639943148918,
+      "grad_norm": 1.30099618434906,
+      "learning_rate": 1.3027941535245191e-06,
+      "loss": 0.174,
+      "step": 58660
+    },
+    {
+      "epoch": 0.772227193767437,
+      "grad_norm": 1.5411639213562012,
+      "learning_rate": 1.299926084474959e-06,
+      "loss": 0.1861,
+      "step": 58680
+    },
+    {
+      "epoch": 0.7724903932199821,
+      "grad_norm": 1.9765437841415405,
+      "learning_rate": 1.297060704116167e-06,
+      "loss": 0.178,
+      "step": 58700
+    },
+    {
+      "epoch": 0.7727535926725272,
+      "grad_norm": 1.0876033306121826,
+      "learning_rate": 1.2941980145302956e-06,
+      "loss": 0.1679,
+      "step": 58720
+    },
+    {
+      "epoch": 0.7730167921250723,
+      "grad_norm": 1.6539831161499023,
+      "learning_rate": 1.2913380177975471e-06,
+      "loss": 0.1928,
+      "step": 58740
+    },
+    {
+      "epoch": 0.7732799915776175,
+      "grad_norm": 1.4020752906799316,
+      "learning_rate": 1.2884807159961582e-06,
+      "loss": 0.1651,
+      "step": 58760
+    },
+    {
+      "epoch": 0.7735431910301627,
+      "grad_norm": 1.3935147523880005,
+      "learning_rate": 1.2856261112024127e-06,
+      "loss": 0.1701,
+      "step": 58780
+    },
+    {
+      "epoch": 0.7738063904827078,
+      "grad_norm": 1.2016308307647705,
+      "learning_rate": 1.2827742054906368e-06,
+      "loss": 0.1704,
+      "step": 58800
+    },
+    {
+      "epoch": 0.774069589935253,
+      "grad_norm": 1.515897274017334,
+      "learning_rate": 1.2799250009331898e-06,
+      "loss": 0.1681,
+      "step": 58820
+    },
+    {
+      "epoch": 0.774332789387798,
+      "grad_norm": 1.5035951137542725,
+      "learning_rate": 1.2770784996004725e-06,
+      "loss": 0.1703,
+      "step": 58840
+    },
+    {
+      "epoch": 0.7745959888403432,
+      "grad_norm": 1.760991096496582,
+      "learning_rate": 1.2742347035609192e-06,
+      "loss": 0.182,
+      "step": 58860
+    },
+    {
+      "epoch": 0.7748591882928884,
+      "grad_norm": 1.6480381488800049,
+      "learning_rate": 1.2713936148809968e-06,
+      "loss": 0.1655,
+      "step": 58880
+    },
+    {
+      "epoch": 0.7751223877454335,
+      "grad_norm": 1.6684755086898804,
+      "learning_rate": 1.2685552356252113e-06,
+      "loss": 0.1721,
+      "step": 58900
+    },
+    {
+      "epoch": 0.7753855871979787,
+      "grad_norm": 2.096904993057251,
+      "learning_rate": 1.265719567856094e-06,
+      "loss": 0.1698,
+      "step": 58920
+    },
+    {
+      "epoch": 0.7756487866505237,
+      "grad_norm": 1.6735893487930298,
+      "learning_rate": 1.262886613634204e-06,
+      "loss": 0.1797,
+      "step": 58940
+    },
+    {
+      "epoch": 0.7759119861030689,
+      "grad_norm": 1.5056802034378052,
+      "learning_rate": 1.260056375018136e-06,
+      "loss": 0.1895,
+      "step": 58960
+    },
+    {
+      "epoch": 0.776175185555614,
+      "grad_norm": 1.2995476722717285,
+      "learning_rate": 1.257228854064505e-06,
+      "loss": 0.1836,
+      "step": 58980
+    },
+    {
+      "epoch": 0.7764383850081592,
+      "grad_norm": 1.375751256942749,
+      "learning_rate": 1.2544040528279577e-06,
+      "loss": 0.1737,
+      "step": 59000
+    },
+    {
+      "epoch": 0.7764383850081592,
+      "eval_loss": 0.201969176530838,
+      "eval_runtime": 4.4472,
+      "eval_samples_per_second": 3.598,
+      "eval_steps_per_second": 1.799,
+      "step": 59000
+    },
+    {
+      "epoch": 0.7767015844607044,
+      "grad_norm": 1.4161776304244995,
+      "learning_rate": 1.2515819733611562e-06,
+      "loss": 0.1878,
+      "step": 59020
+    },
+    {
+      "epoch": 0.7769647839132494,
+      "grad_norm": 1.913475751876831,
+      "learning_rate": 1.2487626177147887e-06,
+      "loss": 0.1815,
+      "step": 59040
+    },
+    {
+      "epoch": 0.7772279833657946,
+      "grad_norm": 1.175351858139038,
+      "learning_rate": 1.2459459879375673e-06,
+      "loss": 0.179,
+      "step": 59060
+    },
+    {
+      "epoch": 0.7774911828183397,
+      "grad_norm": 1.6707308292388916,
+      "learning_rate": 1.2431320860762208e-06,
+      "loss": 0.1684,
+      "step": 59080
+    },
+    {
+      "epoch": 0.7777543822708849,
+      "grad_norm": 1.761462688446045,
+      "learning_rate": 1.2403209141754903e-06,
+      "loss": 0.1587,
+      "step": 59100
+    },
+    {
+      "epoch": 0.77801758172343,
+      "grad_norm": 1.2407562732696533,
+      "learning_rate": 1.2375124742781435e-06,
+      "loss": 0.1711,
+      "step": 59120
+    },
+    {
+      "epoch": 0.7782807811759751,
+      "grad_norm": 1.4488756656646729,
+      "learning_rate": 1.2347067684249537e-06,
+      "loss": 0.1612,
+      "step": 59140
+    },
+    {
+      "epoch": 0.7785439806285203,
+      "grad_norm": 2.037593126296997,
+      "learning_rate": 1.2319037986547167e-06,
+      "loss": 0.1905,
+      "step": 59160
+    },
+    {
+      "epoch": 0.7788071800810654,
+      "grad_norm": 0.9616973400115967,
+      "learning_rate": 1.22910356700423e-06,
+      "loss": 0.165,
+      "step": 59180
+    },
+    {
+      "epoch": 0.7790703795336106,
+      "grad_norm": 1.1789828538894653,
+      "learning_rate": 1.2263060755083057e-06,
+      "loss": 0.1685,
+      "step": 59200
+    },
+    {
+      "epoch": 0.7793335789861557,
+      "grad_norm": 1.1674270629882812,
+      "learning_rate": 1.223511326199769e-06,
+      "loss": 0.1598,
+      "step": 59220
+    },
+    {
+      "epoch": 0.7795967784387009,
+      "grad_norm": 1.8126527070999146,
+      "learning_rate": 1.2207193211094465e-06,
+      "loss": 0.1674,
+      "step": 59240
+    },
+    {
+      "epoch": 0.779859977891246,
+      "grad_norm": 1.95143723487854,
+      "learning_rate": 1.2179300622661733e-06,
+      "loss": 0.1895,
+      "step": 59260
+    },
+    {
+      "epoch": 0.7801231773437911,
+      "grad_norm": 1.8410273790359497,
+      "learning_rate": 1.2151435516967885e-06,
+      "loss": 0.1651,
+      "step": 59280
+    },
+    {
+      "epoch": 0.7803863767963363,
+      "grad_norm": 1.8608765602111816,
+      "learning_rate": 1.2123597914261325e-06,
+      "loss": 0.1645,
+      "step": 59300
+    },
+    {
+      "epoch": 0.7806495762488814,
+      "grad_norm": 1.4851032495498657,
+      "learning_rate": 1.2095787834770522e-06,
+      "loss": 0.1464,
+      "step": 59320
+    },
+    {
+      "epoch": 0.7809127757014266,
+      "grad_norm": 1.4330921173095703,
+      "learning_rate": 1.2068005298703894e-06,
+      "loss": 0.1757,
+      "step": 59340
+    },
+    {
+      "epoch": 0.7811759751539716,
+      "grad_norm": 1.9050829410552979,
+      "learning_rate": 1.2040250326249854e-06,
+      "loss": 0.1773,
+      "step": 59360
+    },
+    {
+      "epoch": 0.7814391746065168,
+      "grad_norm": 1.7833311557769775,
+      "learning_rate": 1.2012522937576804e-06,
+      "loss": 0.1724,
+      "step": 59380
+    },
+    {
+      "epoch": 0.781702374059062,
+      "grad_norm": 1.531018614768982,
+      "learning_rate": 1.1984823152833085e-06,
+      "loss": 0.1793,
+      "step": 59400
+    },
+    {
+      "epoch": 0.7819655735116071,
+      "grad_norm": 2.1883702278137207,
+      "learning_rate": 1.1957150992146992e-06,
+      "loss": 0.1636,
+      "step": 59420
+    },
+    {
+      "epoch": 0.7822287729641523,
+      "grad_norm": 1.9193726778030396,
+      "learning_rate": 1.1929506475626735e-06,
+      "loss": 0.1843,
+      "step": 59440
+    },
+    {
+      "epoch": 0.7824919724166973,
+      "grad_norm": 2.0861921310424805,
+      "learning_rate": 1.1901889623360424e-06,
+      "loss": 0.1693,
+      "step": 59460
+    },
+    {
+      "epoch": 0.7827551718692425,
+      "grad_norm": 1.2736181020736694,
+      "learning_rate": 1.1874300455416116e-06,
+      "loss": 0.1835,
+      "step": 59480
+    },
+    {
+      "epoch": 0.7830183713217876,
+      "grad_norm": 1.4793052673339844,
+      "learning_rate": 1.1846738991841712e-06,
+      "loss": 0.1655,
+      "step": 59500
+    },
+    {
+      "epoch": 0.7830183713217876,
+      "eval_loss": 0.2029777616262436,
+      "eval_runtime": 4.2991,
+      "eval_samples_per_second": 3.722,
+      "eval_steps_per_second": 1.861,
+      "step": 59500
+    },
+    {
+      "epoch": 0.7832815707743328,
+      "grad_norm": 1.7736455202102661,
+      "learning_rate": 1.1819205252664978e-06,
+      "loss": 0.1627,
+      "step": 59520
+    },
+    {
+      "epoch": 0.783544770226878,
+      "grad_norm": 1.3596330881118774,
+      "learning_rate": 1.1791699257893546e-06,
+      "loss": 0.1734,
+      "step": 59540
+    },
+    {
+      "epoch": 0.7838079696794231,
+      "grad_norm": 1.7320374250411987,
+      "learning_rate": 1.1764221027514872e-06,
+      "loss": 0.1712,
+      "step": 59560
+    },
+    {
+      "epoch": 0.7840711691319682,
+      "grad_norm": 1.6869065761566162,
+      "learning_rate": 1.17367705814963e-06,
+      "loss": 0.1512,
+      "step": 59580
+    },
+    {
+      "epoch": 0.7843343685845133,
+      "grad_norm": 1.9573404788970947,
+      "learning_rate": 1.1709347939784865e-06,
+      "loss": 0.1876,
+      "step": 59600
+    },
+    {
+      "epoch": 0.7845975680370585,
+      "grad_norm": 1.758724331855774,
+      "learning_rate": 1.1681953122307521e-06,
+      "loss": 0.1865,
+      "step": 59620
+    },
+    {
+      "epoch": 0.7848607674896037,
+      "grad_norm": 1.8770737648010254,
+      "learning_rate": 1.1654586148970925e-06,
+      "loss": 0.1629,
+      "step": 59640
+    },
+    {
+      "epoch": 0.7851239669421488,
+      "grad_norm": 1.240493655204773,
+      "learning_rate": 1.1627247039661538e-06,
+      "loss": 0.1509,
+      "step": 59660
+    },
+    {
+      "epoch": 0.7853871663946939,
+      "grad_norm": 1.7506701946258545,
+      "learning_rate": 1.1599935814245545e-06,
+      "loss": 0.1627,
+      "step": 59680
+    },
+    {
+      "epoch": 0.785650365847239,
+      "grad_norm": 1.3721647262573242,
+      "learning_rate": 1.1572652492568898e-06,
+      "loss": 0.189,
+      "step": 59700
+    },
+    {
+      "epoch": 0.7859135652997842,
+      "grad_norm": 1.812768816947937,
+      "learning_rate": 1.1545397094457238e-06,
+      "loss": 0.1687,
+      "step": 59720
+    },
+    {
+      "epoch": 0.7861767647523293,
+      "grad_norm": 1.462788462638855,
+      "learning_rate": 1.1518169639715982e-06,
+      "loss": 0.1815,
+      "step": 59740
+    },
+    {
+      "epoch": 0.7864399642048745,
+      "grad_norm": 1.6214158535003662,
+      "learning_rate": 1.1490970148130137e-06,
+      "loss": 0.1609,
+      "step": 59760
+    },
+    {
+      "epoch": 0.7867031636574195,
+      "grad_norm": 2.03948712348938,
+      "learning_rate": 1.1463798639464497e-06,
+      "loss": 0.1834,
+      "step": 59780
+    },
+    {
+      "epoch": 0.7869663631099647,
+      "grad_norm": 1.3401118516921997,
+      "learning_rate": 1.1436655133463449e-06,
+      "loss": 0.1762,
+      "step": 59800
+    },
+    {
+      "epoch": 0.7872295625625099,
+      "grad_norm": 2.065669536590576,
+      "learning_rate": 1.1409539649851066e-06,
+      "loss": 0.177,
+      "step": 59820
+    },
+    {
+      "epoch": 0.787492762015055,
+      "grad_norm": 1.2004154920578003,
+      "learning_rate": 1.1382452208331046e-06,
+      "loss": 0.1585,
+      "step": 59840
+    },
+    {
+      "epoch": 0.7877559614676002,
+      "grad_norm": 1.67708158493042,
+      "learning_rate": 1.1355392828586699e-06,
+      "loss": 0.1888,
+      "step": 59860
+    },
+    {
+      "epoch": 0.7880191609201452,
+      "grad_norm": 1.5968366861343384,
+      "learning_rate": 1.1328361530280986e-06,
+      "loss": 0.1554,
+      "step": 59880
+    },
+    {
+      "epoch": 0.7882823603726904,
+      "grad_norm": 1.5631675720214844,
+      "learning_rate": 1.1301358333056417e-06,
+      "loss": 0.1863,
+      "step": 59900
+    },
+    {
+      "epoch": 0.7885455598252356,
+      "grad_norm": 1.6279438734054565,
+      "learning_rate": 1.127438325653511e-06,
+      "loss": 0.1743,
+      "step": 59920
+    },
+    {
+      "epoch": 0.7888087592777807,
+      "grad_norm": 1.3442350625991821,
+      "learning_rate": 1.1247436320318727e-06,
+      "loss": 0.1742,
+      "step": 59940
+    },
+    {
+      "epoch": 0.7890719587303259,
+      "grad_norm": 2.1185595989227295,
+      "learning_rate": 1.1220517543988495e-06,
+      "loss": 0.17,
+      "step": 59960
+    },
+    {
+      "epoch": 0.789335158182871,
+      "grad_norm": 1.282570242881775,
+      "learning_rate": 1.1193626947105162e-06,
+      "loss": 0.1762,
+      "step": 59980
+    },
+    {
+      "epoch": 0.7895983576354161,
+      "grad_norm": 1.7684134244918823,
+      "learning_rate": 1.1166764549209058e-06,
+      "loss": 0.1617,
+      "step": 60000
+    },
+    {
+      "epoch": 0.7895983576354161,
+      "eval_loss": 0.2026723027229309,
+      "eval_runtime": 4.2789,
+      "eval_samples_per_second": 3.739,
+      "eval_steps_per_second": 1.87,
+      "step": 60000
+    },
+    {
+      "epoch": 0.7898615570879612,
+      "grad_norm": 1.5146923065185547,
+      "learning_rate": 1.1139930369819924e-06,
+      "loss": 0.1682,
+      "step": 60020
+    },
+    {
+      "epoch": 0.7901247565405064,
+      "grad_norm": 1.7564114332199097,
+      "learning_rate": 1.1113124428437082e-06,
+      "loss": 0.1986,
+      "step": 60040
+    },
+    {
+      "epoch": 0.7903879559930516,
+      "grad_norm": 1.5766677856445312,
+      "learning_rate": 1.1086346744539295e-06,
+      "loss": 0.1573,
+      "step": 60060
+    },
+    {
+      "epoch": 0.7906511554455967,
+      "grad_norm": 1.721168875694275,
+      "learning_rate": 1.1059597337584793e-06,
+      "loss": 0.1805,
+      "step": 60080
+    },
+    {
+      "epoch": 0.7909143548981418,
+      "grad_norm": 1.8258212804794312,
+      "learning_rate": 1.103287622701127e-06,
+      "loss": 0.1628,
+      "step": 60100
+    },
+    {
+      "epoch": 0.7911775543506869,
+      "grad_norm": 1.8004575967788696,
+      "learning_rate": 1.1006183432235835e-06,
+      "loss": 0.1675,
+      "step": 60120
+    },
+    {
+      "epoch": 0.7914407538032321,
+      "grad_norm": 1.2917594909667969,
+      "learning_rate": 1.0979518972655067e-06,
+      "loss": 0.1603,
+      "step": 60140
+    },
+    {
+      "epoch": 0.7917039532557772,
+      "grad_norm": 1.6854146718978882,
+      "learning_rate": 1.0952882867644925e-06,
+      "loss": 0.17,
+      "step": 60160
+    },
+    {
+      "epoch": 0.7919671527083224,
+      "grad_norm": 2.6001946926116943,
+      "learning_rate": 1.0926275136560733e-06,
+      "loss": 0.1585,
+      "step": 60180
+    },
+    {
+      "epoch": 0.7922303521608675,
+      "grad_norm": 1.956432580947876,
+      "learning_rate": 1.089969579873726e-06,
+      "loss": 0.1602,
+      "step": 60200
+    },
+    {
+      "epoch": 0.7924935516134126,
+      "grad_norm": 1.7372905015945435,
+      "learning_rate": 1.08731448734886e-06,
+      "loss": 0.1516,
+      "step": 60220
+    },
+    {
+      "epoch": 0.7927567510659578,
+      "grad_norm": 2.0279760360717773,
+      "learning_rate": 1.084662238010822e-06,
+      "loss": 0.176,
+      "step": 60240
+    },
+    {
+      "epoch": 0.7930199505185029,
+      "grad_norm": 1.6804319620132446,
+      "learning_rate": 1.0820128337868907e-06,
+      "loss": 0.1977,
+      "step": 60260
+    },
+    {
+      "epoch": 0.7932831499710481,
+      "grad_norm": 1.3478953838348389,
+      "learning_rate": 1.0793662766022784e-06,
+      "loss": 0.1616,
+      "step": 60280
+    },
+    {
+      "epoch": 0.7935463494235931,
+      "grad_norm": 1.6485393047332764,
+      "learning_rate": 1.0767225683801314e-06,
+      "loss": 0.1611,
+      "step": 60300
+    },
+    {
+      "epoch": 0.7938095488761383,
+      "grad_norm": 1.7671046257019043,
+      "learning_rate": 1.0740817110415214e-06,
+      "loss": 0.1873,
+      "step": 60320
+    },
+    {
+      "epoch": 0.7940727483286835,
+      "grad_norm": 1.1235485076904297,
+      "learning_rate": 1.0714437065054506e-06,
+      "loss": 0.1704,
+      "step": 60340
+    },
+    {
+      "epoch": 0.7943359477812286,
+      "grad_norm": 1.851552963256836,
+      "learning_rate": 1.0688085566888474e-06,
+      "loss": 0.175,
+      "step": 60360
+    },
+    {
+      "epoch": 0.7945991472337738,
+      "grad_norm": 2.0272514820098877,
+      "learning_rate": 1.0661762635065653e-06,
+      "loss": 0.1731,
+      "step": 60380
+    },
+    {
+      "epoch": 0.7948623466863189,
+      "grad_norm": 1.7405811548233032,
+      "learning_rate": 1.0635468288713867e-06,
+      "loss": 0.1709,
+      "step": 60400
+    },
+    {
+      "epoch": 0.795125546138864,
+      "grad_norm": 1.8059755563735962,
+      "learning_rate": 1.0609202546940095e-06,
+      "loss": 0.1682,
+      "step": 60420
+    },
+    {
+      "epoch": 0.7953887455914092,
+      "grad_norm": 1.6559216976165771,
+      "learning_rate": 1.0582965428830567e-06,
+      "loss": 0.188,
+      "step": 60440
+    },
+    {
+      "epoch": 0.7956519450439543,
+      "grad_norm": 2.025222063064575,
+      "learning_rate": 1.0556756953450725e-06,
+      "loss": 0.1701,
+      "step": 60460
+    },
+    {
+      "epoch": 0.7959151444964995,
+      "grad_norm": 1.659895420074463,
+      "learning_rate": 1.0530577139845188e-06,
+      "loss": 0.1642,
+      "step": 60480
+    },
+    {
+      "epoch": 0.7961783439490446,
+      "grad_norm": 1.524987816810608,
+      "learning_rate": 1.0504426007037737e-06,
+      "loss": 0.1667,
+      "step": 60500
+    },
+    {
+      "epoch": 0.7961783439490446,
+      "eval_loss": 0.19712546467781067,
+      "eval_runtime": 4.5404,
+      "eval_samples_per_second": 3.524,
+      "eval_steps_per_second": 1.762,
+      "step": 60500
+    },
+    {
+      "epoch": 0.7964415434015897,
+      "grad_norm": 1.487352728843689,
+      "learning_rate": 1.0478303574031318e-06,
+      "loss": 0.1637,
+      "step": 60520
+    },
+    {
+      "epoch": 0.7967047428541348,
+      "grad_norm": 1.7424488067626953,
+      "learning_rate": 1.0452209859808016e-06,
+      "loss": 0.1733,
+      "step": 60540
+    },
+    {
+      "epoch": 0.79696794230668,
+      "grad_norm": 1.6332670450210571,
+      "learning_rate": 1.0426144883329075e-06,
+      "loss": 0.175,
+      "step": 60560
+    },
+    {
+      "epoch": 0.7972311417592252,
+      "grad_norm": 1.6560280323028564,
+      "learning_rate": 1.0400108663534836e-06,
+      "loss": 0.177,
+      "step": 60580
+    },
+    {
+      "epoch": 0.7974943412117703,
+      "grad_norm": 1.4392493963241577,
+      "learning_rate": 1.0374101219344707e-06,
+      "loss": 0.178,
+      "step": 60600
+    },
+    {
+      "epoch": 0.7977575406643154,
+      "grad_norm": 1.7596055269241333,
+      "learning_rate": 1.034812256965726e-06,
+      "loss": 0.1581,
+      "step": 60620
+    },
+    {
+      "epoch": 0.7980207401168605,
+      "grad_norm": 2.0967814922332764,
+      "learning_rate": 1.0322172733350095e-06,
+      "loss": 0.186,
+      "step": 60640
+    },
+    {
+      "epoch": 0.7982839395694057,
+      "grad_norm": 1.4627220630645752,
+      "learning_rate": 1.0296251729279882e-06,
+      "loss": 0.1722,
+      "step": 60660
+    },
+    {
+      "epoch": 0.7985471390219508,
+      "grad_norm": 1.807447910308838,
+      "learning_rate": 1.0270359576282352e-06,
+      "loss": 0.1791,
+      "step": 60680
+    },
+    {
+      "epoch": 0.798810338474496,
+      "grad_norm": 2.054971694946289,
+      "learning_rate": 1.0244496293172234e-06,
+      "loss": 0.1717,
+      "step": 60700
+    },
+    {
+      "epoch": 0.7990735379270412,
+      "grad_norm": 1.2158349752426147,
+      "learning_rate": 1.021866189874335e-06,
+      "loss": 0.1704,
+      "step": 60720
+    },
+    {
+      "epoch": 0.7993367373795862,
+      "grad_norm": 1.7639482021331787,
+      "learning_rate": 1.019285641176846e-06,
+      "loss": 0.1643,
+      "step": 60740
+    },
+    {
+      "epoch": 0.7995999368321314,
+      "grad_norm": 1.4917110204696655,
+      "learning_rate": 1.016707985099935e-06,
+      "loss": 0.1898,
+      "step": 60760
+    },
+    {
+      "epoch": 0.7998631362846765,
+      "grad_norm": 1.5652217864990234,
+      "learning_rate": 1.0141332235166783e-06,
+      "loss": 0.1705,
+      "step": 60780
+    },
+    {
+      "epoch": 0.8001263357372217,
+      "grad_norm": 1.374100923538208,
+      "learning_rate": 1.0115613582980472e-06,
+      "loss": 0.1695,
+      "step": 60800
+    },
+    {
+      "epoch": 0.8003895351897669,
+      "grad_norm": 1.1942840814590454,
+      "learning_rate": 1.0089923913129135e-06,
+      "loss": 0.1506,
+      "step": 60820
+    },
+    {
+      "epoch": 0.8006527346423119,
+      "grad_norm": 2.26965594291687,
+      "learning_rate": 1.0064263244280358e-06,
+      "loss": 0.1683,
+      "step": 60840
+    },
+    {
+      "epoch": 0.8009159340948571,
+      "grad_norm": 1.337792992591858,
+      "learning_rate": 1.003863159508069e-06,
+      "loss": 0.1718,
+      "step": 60860
+    },
+    {
+      "epoch": 0.8011791335474022,
+      "grad_norm": 1.5901331901550293,
+      "learning_rate": 1.001302898415561e-06,
+      "loss": 0.1699,
+      "step": 60880
+    },
+    {
+      "epoch": 0.8014423329999474,
+      "grad_norm": 1.0626742839813232,
+      "learning_rate": 9.987455430109466e-07,
+      "loss": 0.1791,
+      "step": 60900
+    },
+    {
+      "epoch": 0.8017055324524925,
+      "grad_norm": 1.4814869165420532,
+      "learning_rate": 9.961910951525511e-07,
+      "loss": 0.1779,
+      "step": 60920
+    },
+    {
+      "epoch": 0.8019687319050376,
+      "grad_norm": 1.5241485834121704,
+      "learning_rate": 9.93639556696585e-07,
+      "loss": 0.178,
+      "step": 60940
+    },
+    {
+      "epoch": 0.8022319313575828,
+      "grad_norm": 1.7346006631851196,
+      "learning_rate": 9.910909294971454e-07,
+      "loss": 0.168,
+      "step": 60960
+    },
+    {
+      "epoch": 0.8024951308101279,
+      "grad_norm": 1.786824107170105,
+      "learning_rate": 9.885452154062164e-07,
+      "loss": 0.17,
+      "step": 60980
+    },
+    {
+      "epoch": 0.8027583302626731,
+      "grad_norm": 1.6639118194580078,
+      "learning_rate": 9.86002416273662e-07,
+      "loss": 0.1634,
+      "step": 61000
+    },
+    {
+      "epoch": 0.8027583302626731,
+      "eval_loss": 0.20073607563972473,
+      "eval_runtime": 3.9878,
+      "eval_samples_per_second": 4.012,
+      "eval_steps_per_second": 2.006,
+      "step": 61000
+    },
+    {
+      "epoch": 0.8030215297152182,
+      "grad_norm": 1.2318739891052246,
+      "learning_rate": 9.834625339472299e-07,
+      "loss": 0.1689,
+      "step": 61020
+    },
+    {
+      "epoch": 0.8032847291677633,
+      "grad_norm": 1.368556022644043,
+      "learning_rate": 9.809255702725462e-07,
+      "loss": 0.1762,
+      "step": 61040
+    },
+    {
+      "epoch": 0.8035479286203084,
+      "grad_norm": 1.433231234550476,
+      "learning_rate": 9.783915270931189e-07,
+      "loss": 0.1602,
+      "step": 61060
+    },
+    {
+      "epoch": 0.8038111280728536,
+      "grad_norm": 2.020117998123169,
+      "learning_rate": 9.75860406250332e-07,
+      "loss": 0.1603,
+      "step": 61080
+    },
+    {
+      "epoch": 0.8040743275253988,
+      "grad_norm": 1.1014788150787354,
+      "learning_rate": 9.733322095834457e-07,
+      "loss": 0.1591,
+      "step": 61100
+    },
+    {
+      "epoch": 0.8043375269779439,
+      "grad_norm": 1.6257692575454712,
+      "learning_rate": 9.70806938929596e-07,
+      "loss": 0.1595,
+      "step": 61120
+    },
+    {
+      "epoch": 0.8046007264304891,
+      "grad_norm": 1.2954622507095337,
+      "learning_rate": 9.68284596123794e-07,
+      "loss": 0.1568,
+      "step": 61140
+    },
+    {
+      "epoch": 0.8048639258830341,
+      "grad_norm": 1.7231395244598389,
+      "learning_rate": 9.657651829989212e-07,
+      "loss": 0.1677,
+      "step": 61160
+    },
+    {
+      "epoch": 0.8051271253355793,
+      "grad_norm": 1.4518773555755615,
+      "learning_rate": 9.632487013857317e-07,
+      "loss": 0.195,
+      "step": 61180
+    },
+    {
+      "epoch": 0.8053903247881244,
+      "grad_norm": 1.9091700315475464,
+      "learning_rate": 9.607351531128473e-07,
+      "loss": 0.1833,
+      "step": 61200
+    },
+    {
+      "epoch": 0.8056535242406696,
+      "grad_norm": 2.0551693439483643,
+      "learning_rate": 9.582245400067592e-07,
+      "loss": 0.1837,
+      "step": 61220
+    },
+    {
+      "epoch": 0.8059167236932148,
+      "grad_norm": 1.934097170829773,
+      "learning_rate": 9.5571686389183e-07,
+      "loss": 0.1783,
+      "step": 61240
+    },
+    {
+      "epoch": 0.8061799231457598,
+      "grad_norm": 1.960164189338684,
+      "learning_rate": 9.532121265902788e-07,
+      "loss": 0.176,
+      "step": 61260
+    },
+    {
+      "epoch": 0.806443122598305,
+      "grad_norm": 1.476594090461731,
+      "learning_rate": 9.507103299221987e-07,
+      "loss": 0.1727,
+      "step": 61280
+    },
+    {
+      "epoch": 0.8067063220508501,
+      "grad_norm": 1.3404552936553955,
+      "learning_rate": 9.482114757055411e-07,
+      "loss": 0.1593,
+      "step": 61300
+    },
+    {
+      "epoch": 0.8069695215033953,
+      "grad_norm": 1.5193054676055908,
+      "learning_rate": 9.45715565756119e-07,
+      "loss": 0.1765,
+      "step": 61320
+    },
+    {
+      "epoch": 0.8072327209559405,
+      "grad_norm": 1.2150378227233887,
+      "learning_rate": 9.432226018876078e-07,
+      "loss": 0.148,
+      "step": 61340
+    },
+    {
+      "epoch": 0.8074959204084855,
+      "grad_norm": 1.625303864479065,
+      "learning_rate": 9.407325859115413e-07,
+      "loss": 0.169,
+      "step": 61360
+    },
+    {
+      "epoch": 0.8077591198610307,
+      "grad_norm": 1.5935755968093872,
+      "learning_rate": 9.382455196373091e-07,
+      "loss": 0.1835,
+      "step": 61380
+    },
+    {
+      "epoch": 0.8080223193135758,
+      "grad_norm": 2.030334949493408,
+      "learning_rate": 9.357614048721647e-07,
+      "loss": 0.1699,
+      "step": 61400
+    },
+    {
+      "epoch": 0.808285518766121,
+      "grad_norm": 1.967953085899353,
+      "learning_rate": 9.332802434212052e-07,
+      "loss": 0.1852,
+      "step": 61420
+    },
+    {
+      "epoch": 0.8085487182186661,
+      "grad_norm": 1.5830293893814087,
+      "learning_rate": 9.308020370873921e-07,
+      "loss": 0.1702,
+      "step": 61440
+    },
+    {
+      "epoch": 0.8088119176712112,
+      "grad_norm": 1.4843531847000122,
+      "learning_rate": 9.283267876715352e-07,
+      "loss": 0.1695,
+      "step": 61460
+    },
+    {
+      "epoch": 0.8090751171237563,
+      "grad_norm": 1.98789644241333,
+      "learning_rate": 9.258544969722949e-07,
+      "loss": 0.1579,
+      "step": 61480
+    },
+    {
+      "epoch": 0.8093383165763015,
+      "grad_norm": 2.0813450813293457,
+      "learning_rate": 9.233851667861832e-07,
+      "loss": 0.1693,
+      "step": 61500
+    },
+    {
+      "epoch": 0.8093383165763015,
+      "eval_loss": 0.20448192954063416,
+      "eval_runtime": 4.2982,
+      "eval_samples_per_second": 3.722,
+      "eval_steps_per_second": 1.861,
+      "step": 61500
+    },
+    {
+      "epoch": 0.8096015160288467,
+      "grad_norm": 1.1709486246109009,
+      "learning_rate": 9.209187989075585e-07,
+      "loss": 0.1492,
+      "step": 61520
+    },
+    {
+      "epoch": 0.8098647154813918,
+      "grad_norm": 1.894989013671875,
+      "learning_rate": 9.184553951286312e-07,
+      "loss": 0.1834,
+      "step": 61540
+    },
+    {
+      "epoch": 0.810127914933937,
+      "grad_norm": 1.3323801755905151,
+      "learning_rate": 9.159949572394539e-07,
+      "loss": 0.1858,
+      "step": 61560
+    },
+    {
+      "epoch": 0.810391114386482,
+      "grad_norm": 2.1492772102355957,
+      "learning_rate": 9.135374870279251e-07,
+      "loss": 0.1687,
+      "step": 61580
+    },
+    {
+      "epoch": 0.8106543138390272,
+      "grad_norm": 1.0841501951217651,
+      "learning_rate": 9.110829862797865e-07,
+      "loss": 0.1633,
+      "step": 61600
+    },
+    {
+      "epoch": 0.8109175132915724,
+      "grad_norm": 1.6370728015899658,
+      "learning_rate": 9.086314567786225e-07,
+      "loss": 0.1644,
+      "step": 61620
+    },
+    {
+      "epoch": 0.8111807127441175,
+      "grad_norm": 1.655275583267212,
+      "learning_rate": 9.061829003058576e-07,
+      "loss": 0.1732,
+      "step": 61640
+    },
+    {
+      "epoch": 0.8114439121966627,
+      "grad_norm": 1.3513749837875366,
+      "learning_rate": 9.037373186407594e-07,
+      "loss": 0.1736,
+      "step": 61660
+    },
+    {
+      "epoch": 0.8117071116492077,
+      "grad_norm": 2.0674614906311035,
+      "learning_rate": 9.01294713560426e-07,
+      "loss": 0.1666,
+      "step": 61680
+    },
+    {
+      "epoch": 0.8119703111017529,
+      "grad_norm": 1.7818069458007812,
+      "learning_rate": 8.988550868398021e-07,
+      "loss": 0.1706,
+      "step": 61700
+    },
+    {
+      "epoch": 0.812233510554298,
+      "grad_norm": 1.224810242652893,
+      "learning_rate": 8.964184402516618e-07,
+      "loss": 0.192,
+      "step": 61720
+    },
+    {
+      "epoch": 0.8124967100068432,
+      "grad_norm": 1.7139692306518555,
+      "learning_rate": 8.939847755666154e-07,
+      "loss": 0.1732,
+      "step": 61740
+    },
+    {
+      "epoch": 0.8127599094593884,
+      "grad_norm": 2.027759313583374,
+      "learning_rate": 8.915540945531065e-07,
+      "loss": 0.1726,
+      "step": 61760
+    },
+    {
+      "epoch": 0.8130231089119334,
+      "grad_norm": 1.502548336982727,
+      "learning_rate": 8.891263989774096e-07,
+      "loss": 0.1775,
+      "step": 61780
+    },
+    {
+      "epoch": 0.8132863083644786,
+      "grad_norm": 1.560992956161499,
+      "learning_rate": 8.86701690603633e-07,
+      "loss": 0.1634,
+      "step": 61800
+    },
+    {
+      "epoch": 0.8135495078170237,
+      "grad_norm": 1.5679430961608887,
+      "learning_rate": 8.842799711937117e-07,
+      "loss": 0.1749,
+      "step": 61820
+    },
+    {
+      "epoch": 0.8138127072695689,
+      "grad_norm": 1.943096399307251,
+      "learning_rate": 8.818612425074064e-07,
+      "loss": 0.1605,
+      "step": 61840
+    },
+    {
+      "epoch": 0.814075906722114,
+      "grad_norm": 1.5279203653335571,
+      "learning_rate": 8.794455063023095e-07,
+      "loss": 0.1667,
+      "step": 61860
+    },
+    {
+      "epoch": 0.8143391061746592,
+      "grad_norm": 1.768259882926941,
+      "learning_rate": 8.77032764333835e-07,
+      "loss": 0.2104,
+      "step": 61880
+    },
+    {
+      "epoch": 0.8146023056272043,
+      "grad_norm": 1.7578437328338623,
+      "learning_rate": 8.746230183552268e-07,
+      "loss": 0.182,
+      "step": 61900
+    },
+    {
+      "epoch": 0.8148655050797494,
+      "grad_norm": 2.163228988647461,
+      "learning_rate": 8.722162701175435e-07,
+      "loss": 0.1627,
+      "step": 61920
+    },
+    {
+      "epoch": 0.8151287045322946,
+      "grad_norm": 1.2159861326217651,
+      "learning_rate": 8.698125213696695e-07,
+      "loss": 0.1618,
+      "step": 61940
+    },
+    {
+      "epoch": 0.8153919039848397,
+      "grad_norm": 1.426085352897644,
+      "learning_rate": 8.674117738583121e-07,
+      "loss": 0.1778,
+      "step": 61960
+    },
+    {
+      "epoch": 0.8156551034373849,
+      "grad_norm": 2.039252519607544,
+      "learning_rate": 8.650140293279946e-07,
+      "loss": 0.1774,
+      "step": 61980
+    },
+    {
+      "epoch": 0.81591830288993,
+      "grad_norm": 1.4729294776916504,
+      "learning_rate": 8.626192895210578e-07,
+      "loss": 0.1646,
+      "step": 62000
+    },
+    {
+      "epoch": 0.81591830288993,
+      "eval_loss": 0.20360016822814941,
+      "eval_runtime": 3.5957,
+      "eval_samples_per_second": 4.45,
+      "eval_steps_per_second": 2.225,
+      "step": 62000
+    },
+    {
+      "epoch": 0.8161815023424751,
+      "grad_norm": 1.2383685111999512,
+      "learning_rate": 8.602275561776613e-07,
+      "loss": 0.1574,
+      "step": 62020
+    },
+    {
+      "epoch": 0.8164447017950203,
+      "grad_norm": 2.248575448989868,
+      "learning_rate": 8.578388310357772e-07,
+      "loss": 0.1678,
+      "step": 62040
+    },
+    {
+      "epoch": 0.8167079012475654,
+      "grad_norm": 1.7336132526397705,
+      "learning_rate": 8.554531158311968e-07,
+      "loss": 0.1727,
+      "step": 62060
+    },
+    {
+      "epoch": 0.8169711007001106,
+      "grad_norm": 1.6267207860946655,
+      "learning_rate": 8.53070412297517e-07,
+      "loss": 0.1695,
+      "step": 62080
+    },
+    {
+      "epoch": 0.8172343001526556,
+      "grad_norm": 1.9440081119537354,
+      "learning_rate": 8.506907221661498e-07,
+      "loss": 0.1755,
+      "step": 62100
+    },
+    {
+      "epoch": 0.8174974996052008,
+      "grad_norm": 2.0126092433929443,
+      "learning_rate": 8.48314047166321e-07,
+      "loss": 0.1742,
+      "step": 62120
+    },
+    {
+      "epoch": 0.817760699057746,
+      "grad_norm": 1.9106305837631226,
+      "learning_rate": 8.4594038902506e-07,
+      "loss": 0.1667,
+      "step": 62140
+    },
+    {
+      "epoch": 0.8180238985102911,
+      "grad_norm": 1.7019164562225342,
+      "learning_rate": 8.435697494672062e-07,
+      "loss": 0.167,
+      "step": 62160
+    },
+    {
+      "epoch": 0.8182870979628363,
+      "grad_norm": 1.2636555433273315,
+      "learning_rate": 8.412021302154055e-07,
+      "loss": 0.1495,
+      "step": 62180
+    },
+    {
+      "epoch": 0.8185502974153813,
+      "grad_norm": 1.535785436630249,
+      "learning_rate": 8.388375329901083e-07,
+      "loss": 0.158,
+      "step": 62200
+    },
+    {
+      "epoch": 0.8188134968679265,
+      "grad_norm": 1.4260268211364746,
+      "learning_rate": 8.364759595095712e-07,
+      "loss": 0.1767,
+      "step": 62220
+    },
+    {
+      "epoch": 0.8190766963204716,
+      "grad_norm": 1.7681612968444824,
+      "learning_rate": 8.34117411489852e-07,
+      "loss": 0.1616,
+      "step": 62240
+    },
+    {
+      "epoch": 0.8193398957730168,
+      "grad_norm": 1.2298855781555176,
+      "learning_rate": 8.317618906448072e-07,
+      "loss": 0.1769,
+      "step": 62260
+    },
+    {
+      "epoch": 0.819603095225562,
+      "grad_norm": 1.7257241010665894,
+      "learning_rate": 8.294093986860991e-07,
+      "loss": 0.1706,
+      "step": 62280
+    },
+    {
+      "epoch": 0.8198662946781071,
+      "grad_norm": 1.7841325998306274,
+      "learning_rate": 8.270599373231863e-07,
+      "loss": 0.1853,
+      "step": 62300
+    },
+    {
+      "epoch": 0.8201294941306522,
+      "grad_norm": 1.5362474918365479,
+      "learning_rate": 8.24713508263324e-07,
+      "loss": 0.1616,
+      "step": 62320
+    },
+    {
+      "epoch": 0.8203926935831973,
+      "grad_norm": 1.8526208400726318,
+      "learning_rate": 8.223701132115663e-07,
+      "loss": 0.1882,
+      "step": 62340
+    },
+    {
+      "epoch": 0.8206558930357425,
+      "grad_norm": 1.7994680404663086,
+      "learning_rate": 8.200297538707596e-07,
+      "loss": 0.181,
+      "step": 62360
+    },
+    {
+      "epoch": 0.8209190924882876,
+      "grad_norm": 1.6399792432785034,
+      "learning_rate": 8.176924319415491e-07,
+      "loss": 0.1649,
+      "step": 62380
+    },
+    {
+      "epoch": 0.8211822919408328,
+      "grad_norm": 1.714515209197998,
+      "learning_rate": 8.153581491223689e-07,
+      "loss": 0.1516,
+      "step": 62400
+    },
+    {
+      "epoch": 0.8214454913933779,
+      "grad_norm": 1.8358033895492554,
+      "learning_rate": 8.130269071094454e-07,
+      "loss": 0.1703,
+      "step": 62420
+    },
+    {
+      "epoch": 0.821708690845923,
+      "grad_norm": 1.2653934955596924,
+      "learning_rate": 8.106987075967959e-07,
+      "loss": 0.1657,
+      "step": 62440
+    },
+    {
+      "epoch": 0.8219718902984682,
+      "grad_norm": 2.2223899364471436,
+      "learning_rate": 8.08373552276226e-07,
+      "loss": 0.1761,
+      "step": 62460
+    },
+    {
+      "epoch": 0.8222350897510133,
+      "grad_norm": 1.365706205368042,
+      "learning_rate": 8.060514428373329e-07,
+      "loss": 0.1744,
+      "step": 62480
+    },
+    {
+      "epoch": 0.8224982892035585,
+      "grad_norm": 1.7462044954299927,
+      "learning_rate": 8.037323809674947e-07,
+      "loss": 0.1625,
+      "step": 62500
+    },
+    {
+      "epoch": 0.8224982892035585,
+      "eval_loss": 0.21124449372291565,
+      "eval_runtime": 4.4412,
+      "eval_samples_per_second": 3.603,
+      "eval_steps_per_second": 1.801,
+      "step": 62500
+    },
+    {
+      "epoch": 0.8227614886561035,
+      "grad_norm": 2.0483758449554443,
+      "learning_rate": 8.014163683518766e-07,
+      "loss": 0.1649,
+      "step": 62520
+    },
+    {
+      "epoch": 0.8230246881086487,
+      "grad_norm": 1.894225001335144,
+      "learning_rate": 7.991034066734332e-07,
+      "loss": 0.1793,
+      "step": 62540
+    },
+    {
+      "epoch": 0.8232878875611939,
+      "grad_norm": 1.8974770307540894,
+      "learning_rate": 7.967934976128955e-07,
+      "loss": 0.1752,
+      "step": 62560
+    },
+    {
+      "epoch": 0.823551087013739,
+      "grad_norm": 1.7720290422439575,
+      "learning_rate": 7.944866428487802e-07,
+      "loss": 0.1739,
+      "step": 62580
+    },
+    {
+      "epoch": 0.8238142864662842,
+      "grad_norm": 1.6768693923950195,
+      "learning_rate": 7.92182844057382e-07,
+      "loss": 0.1556,
+      "step": 62600
+    },
+    {
+      "epoch": 0.8240774859188292,
+      "grad_norm": 1.7112298011779785,
+      "learning_rate": 7.898821029127762e-07,
+      "loss": 0.1515,
+      "step": 62620
+    },
+    {
+      "epoch": 0.8243406853713744,
+      "grad_norm": 2.433269739151001,
+      "learning_rate": 7.875844210868183e-07,
+      "loss": 0.1718,
+      "step": 62640
+    },
+    {
+      "epoch": 0.8246038848239196,
+      "grad_norm": 2.194413900375366,
+      "learning_rate": 7.852898002491371e-07,
+      "loss": 0.1664,
+      "step": 62660
+    },
+    {
+      "epoch": 0.8248670842764647,
+      "grad_norm": 1.6503047943115234,
+      "learning_rate": 7.829982420671395e-07,
+      "loss": 0.1697,
+      "step": 62680
+    },
+    {
+      "epoch": 0.8251302837290099,
+      "grad_norm": 2.2997729778289795,
+      "learning_rate": 7.80709748206005e-07,
+      "loss": 0.155,
+      "step": 62700
+    },
+    {
+      "epoch": 0.825393483181555,
+      "grad_norm": 1.7964937686920166,
+      "learning_rate": 7.784243203286884e-07,
+      "loss": 0.1704,
+      "step": 62720
+    },
+    {
+      "epoch": 0.8256566826341001,
+      "grad_norm": 1.3904101848602295,
+      "learning_rate": 7.761419600959158e-07,
+      "loss": 0.1577,
+      "step": 62740
+    },
+    {
+      "epoch": 0.8259198820866452,
+      "grad_norm": 1.3406258821487427,
+      "learning_rate": 7.738626691661832e-07,
+      "loss": 0.1499,
+      "step": 62760
+    },
+    {
+      "epoch": 0.8261830815391904,
+      "grad_norm": 1.8296293020248413,
+      "learning_rate": 7.715864491957564e-07,
+      "loss": 0.1587,
+      "step": 62780
+    },
+    {
+      "epoch": 0.8264462809917356,
+      "grad_norm": 1.6669952869415283,
+      "learning_rate": 7.693133018386728e-07,
+      "loss": 0.1827,
+      "step": 62800
+    },
+    {
+      "epoch": 0.8267094804442807,
+      "grad_norm": 1.2716376781463623,
+      "learning_rate": 7.670432287467333e-07,
+      "loss": 0.169,
+      "step": 62820
+    },
+    {
+      "epoch": 0.8269726798968258,
+      "grad_norm": 2.1493420600891113,
+      "learning_rate": 7.647762315695062e-07,
+      "loss": 0.1775,
+      "step": 62840
+    },
+    {
+      "epoch": 0.8272358793493709,
+      "grad_norm": 1.6719298362731934,
+      "learning_rate": 7.625123119543249e-07,
+      "loss": 0.1637,
+      "step": 62860
+    },
+    {
+      "epoch": 0.8274990788019161,
+      "grad_norm": 1.8326077461242676,
+      "learning_rate": 7.602514715462845e-07,
+      "loss": 0.1506,
+      "step": 62880
+    },
+    {
+      "epoch": 0.8277622782544612,
+      "grad_norm": 1.3685365915298462,
+      "learning_rate": 7.57993711988248e-07,
+      "loss": 0.1602,
+      "step": 62900
+    },
+    {
+      "epoch": 0.8280254777070064,
+      "grad_norm": 1.429232120513916,
+      "learning_rate": 7.55739034920831e-07,
+      "loss": 0.1667,
+      "step": 62920
+    },
+    {
+      "epoch": 0.8282886771595515,
+      "grad_norm": 1.7953139543533325,
+      "learning_rate": 7.534874419824184e-07,
+      "loss": 0.1703,
+      "step": 62940
+    },
+    {
+      "epoch": 0.8285518766120966,
+      "grad_norm": 1.599197506904602,
+      "learning_rate": 7.512389348091481e-07,
+      "loss": 0.1844,
+      "step": 62960
+    },
+    {
+      "epoch": 0.8288150760646418,
+      "grad_norm": 2.0370492935180664,
+      "learning_rate": 7.489935150349176e-07,
+      "loss": 0.1716,
+      "step": 62980
+    },
+    {
+      "epoch": 0.8290782755171869,
+      "grad_norm": 2.254634380340576,
+      "learning_rate": 7.467511842913799e-07,
+      "loss": 0.1489,
+      "step": 63000
+    },
+    {
+      "epoch": 0.8290782755171869,
+      "eval_loss": 0.19870787858963013,
+      "eval_runtime": 4.3998,
+      "eval_samples_per_second": 3.637,
+      "eval_steps_per_second": 1.818,
+      "step": 63000
+    },
+    {
+      "epoch": 0.8293414749697321,
+      "grad_norm": 1.6619278192520142,
+      "learning_rate": 7.445119442079452e-07,
+      "loss": 0.1662,
+      "step": 63020
+    },
+    {
+      "epoch": 0.8296046744222773,
+      "grad_norm": 2.093893527984619,
+      "learning_rate": 7.422757964117744e-07,
+      "loss": 0.1746,
+      "step": 63040
+    },
+    {
+      "epoch": 0.8298678738748223,
+      "grad_norm": 1.803030252456665,
+      "learning_rate": 7.400427425277873e-07,
+      "loss": 0.1673,
+      "step": 63060
+    },
+    {
+      "epoch": 0.8301310733273675,
+      "grad_norm": 1.6091265678405762,
+      "learning_rate": 7.378127841786481e-07,
+      "loss": 0.1873,
+      "step": 63080
+    },
+    {
+      "epoch": 0.8303942727799126,
+      "grad_norm": 2.215230703353882,
+      "learning_rate": 7.355859229847778e-07,
+      "loss": 0.1558,
+      "step": 63100
+    },
+    {
+      "epoch": 0.8306574722324578,
+      "grad_norm": 2.0143656730651855,
+      "learning_rate": 7.333621605643437e-07,
+      "loss": 0.1631,
+      "step": 63120
+    },
+    {
+      "epoch": 0.8309206716850029,
+      "grad_norm": 1.5690977573394775,
+      "learning_rate": 7.311414985332627e-07,
+      "loss": 0.1575,
+      "step": 63140
+    },
+    {
+      "epoch": 0.831183871137548,
+      "grad_norm": 1.8018003702163696,
+      "learning_rate": 7.289239385051966e-07,
+      "loss": 0.166,
+      "step": 63160
+    },
+    {
+      "epoch": 0.8314470705900932,
+      "grad_norm": 1.5751218795776367,
+      "learning_rate": 7.267094820915549e-07,
+      "loss": 0.1622,
+      "step": 63180
+    },
+    {
+      "epoch": 0.8317102700426383,
+      "grad_norm": 1.7388074398040771,
+      "learning_rate": 7.244981309014926e-07,
+      "loss": 0.1696,
+      "step": 63200
+    },
+    {
+      "epoch": 0.8319734694951835,
+      "grad_norm": 1.595644235610962,
+      "learning_rate": 7.22289886541907e-07,
+      "loss": 0.165,
+      "step": 63220
+    },
+    {
+      "epoch": 0.8322366689477286,
+      "grad_norm": 1.6749718189239502,
+      "learning_rate": 7.200847506174379e-07,
+      "loss": 0.1617,
+      "step": 63240
+    },
+    {
+      "epoch": 0.8324998684002737,
+      "grad_norm": 2.148689031600952,
+      "learning_rate": 7.178827247304659e-07,
+      "loss": 0.1881,
+      "step": 63260
+    },
+    {
+      "epoch": 0.8327630678528188,
+      "grad_norm": 1.331956386566162,
+      "learning_rate": 7.156838104811114e-07,
+      "loss": 0.1856,
+      "step": 63280
+    },
+    {
+      "epoch": 0.833026267305364,
+      "grad_norm": 1.4940193891525269,
+      "learning_rate": 7.13488009467237e-07,
+      "loss": 0.1619,
+      "step": 63300
+    },
+    {
+      "epoch": 0.8332894667579092,
+      "grad_norm": 1.8249919414520264,
+      "learning_rate": 7.1129532328444e-07,
+      "loss": 0.1641,
+      "step": 63320
+    },
+    {
+      "epoch": 0.8335526662104543,
+      "grad_norm": 1.48263418674469,
+      "learning_rate": 7.091057535260515e-07,
+      "loss": 0.182,
+      "step": 63340
+    },
+    {
+      "epoch": 0.8338158656629994,
+      "grad_norm": 1.2354888916015625,
+      "learning_rate": 7.069193017831449e-07,
+      "loss": 0.1623,
+      "step": 63360
+    },
+    {
+      "epoch": 0.8340790651155445,
+      "grad_norm": 1.458530306816101,
+      "learning_rate": 7.047359696445233e-07,
+      "loss": 0.1663,
+      "step": 63380
+    },
+    {
+      "epoch": 0.8343422645680897,
+      "grad_norm": 1.7563568353652954,
+      "learning_rate": 7.025557586967241e-07,
+      "loss": 0.1532,
+      "step": 63400
+    },
+    {
+      "epoch": 0.8346054640206348,
+      "grad_norm": 1.76023268699646,
+      "learning_rate": 7.003786705240157e-07,
+      "loss": 0.1641,
+      "step": 63420
+    },
+    {
+      "epoch": 0.83486866347318,
+      "grad_norm": 1.5677099227905273,
+      "learning_rate": 6.982047067083974e-07,
+      "loss": 0.1695,
+      "step": 63440
+    },
+    {
+      "epoch": 0.8351318629257252,
+      "grad_norm": 1.2566893100738525,
+      "learning_rate": 6.960338688296009e-07,
+      "loss": 0.1679,
+      "step": 63460
+    },
+    {
+      "epoch": 0.8353950623782702,
+      "grad_norm": 2.1114273071289062,
+      "learning_rate": 6.938661584650852e-07,
+      "loss": 0.1612,
+      "step": 63480
+    },
+    {
+      "epoch": 0.8356582618308154,
+      "grad_norm": 1.628509759902954,
+      "learning_rate": 6.917015771900315e-07,
+      "loss": 0.1687,
+      "step": 63500
+    },
+    {
+      "epoch": 0.8356582618308154,
+      "eval_loss": 0.19609299302101135,
+      "eval_runtime": 4.3012,
+      "eval_samples_per_second": 3.72,
+      "eval_steps_per_second": 1.86,
+      "step": 63500
+    },
+    {
+      "epoch": 0.8359214612833605,
+      "grad_norm": 2.2502388954162598,
+      "learning_rate": 6.895401265773549e-07,
+      "loss": 0.169,
+      "step": 63520
+    },
+    {
+      "epoch": 0.8361846607359057,
+      "grad_norm": 2.2081587314605713,
+      "learning_rate": 6.873818081976908e-07,
+      "loss": 0.1587,
+      "step": 63540
+    },
+    {
+      "epoch": 0.8364478601884509,
+      "grad_norm": 1.3546804189682007,
+      "learning_rate": 6.852266236194022e-07,
+      "loss": 0.1792,
+      "step": 63560
+    },
+    {
+      "epoch": 0.8367110596409959,
+      "grad_norm": 1.5917916297912598,
+      "learning_rate": 6.830745744085698e-07,
+      "loss": 0.1599,
+      "step": 63580
+    },
+    {
+      "epoch": 0.8369742590935411,
+      "grad_norm": 1.2565191984176636,
+      "learning_rate": 6.809256621289994e-07,
+      "loss": 0.1769,
+      "step": 63600
+    },
+    {
+      "epoch": 0.8372374585460862,
+      "grad_norm": 1.5627551078796387,
+      "learning_rate": 6.787798883422186e-07,
+      "loss": 0.1763,
+      "step": 63620
+    },
+    {
+      "epoch": 0.8375006579986314,
+      "grad_norm": 1.595245599746704,
+      "learning_rate": 6.766372546074707e-07,
+      "loss": 0.1798,
+      "step": 63640
+    },
+    {
+      "epoch": 0.8377638574511765,
+      "grad_norm": 4.820520401000977,
+      "learning_rate": 6.7449776248172e-07,
+      "loss": 0.1808,
+      "step": 63660
+    },
+    {
+      "epoch": 0.8380270569037216,
+      "grad_norm": 1.8583229780197144,
+      "learning_rate": 6.72361413519646e-07,
+      "loss": 0.184,
+      "step": 63680
+    },
+    {
+      "epoch": 0.8382902563562667,
+      "grad_norm": 1.8213855028152466,
+      "learning_rate": 6.702282092736451e-07,
+      "loss": 0.1823,
+      "step": 63700
+    },
+    {
+      "epoch": 0.8385534558088119,
+      "grad_norm": 1.5587085485458374,
+      "learning_rate": 6.680981512938317e-07,
+      "loss": 0.1711,
+      "step": 63720
+    },
+    {
+      "epoch": 0.8388166552613571,
+      "grad_norm": 1.561562418937683,
+      "learning_rate": 6.659712411280273e-07,
+      "loss": 0.1661,
+      "step": 63740
+    },
+    {
+      "epoch": 0.8390798547139022,
+      "grad_norm": 1.5877907276153564,
+      "learning_rate": 6.638474803217693e-07,
+      "loss": 0.1703,
+      "step": 63760
+    },
+    {
+      "epoch": 0.8393430541664474,
+      "grad_norm": 1.4619585275650024,
+      "learning_rate": 6.6172687041831e-07,
+      "loss": 0.1651,
+      "step": 63780
+    },
+    {
+      "epoch": 0.8396062536189924,
+      "grad_norm": 1.6107808351516724,
+      "learning_rate": 6.596094129586073e-07,
+      "loss": 0.1848,
+      "step": 63800
+    },
+    {
+      "epoch": 0.8398694530715376,
+      "grad_norm": 1.3837153911590576,
+      "learning_rate": 6.574951094813298e-07,
+      "loss": 0.1782,
+      "step": 63820
+    },
+    {
+      "epoch": 0.8401326525240828,
+      "grad_norm": 1.4803657531738281,
+      "learning_rate": 6.553839615228546e-07,
+      "loss": 0.1708,
+      "step": 63840
+    },
+    {
+      "epoch": 0.8403958519766279,
+      "grad_norm": 1.4716441631317139,
+      "learning_rate": 6.532759706172648e-07,
+      "loss": 0.1754,
+      "step": 63860
+    },
+    {
+      "epoch": 0.8406590514291731,
+      "grad_norm": 1.7880501747131348,
+      "learning_rate": 6.511711382963526e-07,
+      "loss": 0.1766,
+      "step": 63880
+    },
+    {
+      "epoch": 0.8409222508817181,
+      "grad_norm": 2.1928865909576416,
+      "learning_rate": 6.490694660896119e-07,
+      "loss": 0.1841,
+      "step": 63900
+    },
+    {
+      "epoch": 0.8411854503342633,
+      "grad_norm": 1.6590133905410767,
+      "learning_rate": 6.469709555242387e-07,
+      "loss": 0.1719,
+      "step": 63920
+    },
+    {
+      "epoch": 0.8414486497868084,
+      "grad_norm": 1.448829174041748,
+      "learning_rate": 6.448756081251362e-07,
+      "loss": 0.1542,
+      "step": 63940
+    },
+    {
+      "epoch": 0.8417118492393536,
+      "grad_norm": 1.0552313327789307,
+      "learning_rate": 6.427834254149062e-07,
+      "loss": 0.1742,
+      "step": 63960
+    },
+    {
+      "epoch": 0.8419750486918988,
+      "grad_norm": 1.4594645500183105,
+      "learning_rate": 6.406944089138517e-07,
+      "loss": 0.1616,
+      "step": 63980
+    },
+    {
+      "epoch": 0.8422382481444438,
+      "grad_norm": 1.6301876306533813,
+      "learning_rate": 6.386085601399738e-07,
+      "loss": 0.1666,
+      "step": 64000
+    },
+    {
+      "epoch": 0.8422382481444438,
+      "eval_loss": 0.20485447347164154,
+      "eval_runtime": 4.8383,
+      "eval_samples_per_second": 3.307,
+      "eval_steps_per_second": 1.653,
+      "step": 64000
+    },
+    {
+      "epoch": 0.842501447596989,
+      "grad_norm": 1.8762991428375244,
+      "learning_rate": 6.365258806089719e-07,
+      "loss": 0.1689,
+      "step": 64020
+    },
+    {
+      "epoch": 0.8427646470495341,
+      "grad_norm": 1.4581221342086792,
+      "learning_rate": 6.344463718342458e-07,
+      "loss": 0.1863,
+      "step": 64040
+    },
+    {
+      "epoch": 0.8430278465020793,
+      "grad_norm": 1.8183478116989136,
+      "learning_rate": 6.323700353268863e-07,
+      "loss": 0.1683,
+      "step": 64060
+    },
+    {
+      "epoch": 0.8432910459546245,
+      "grad_norm": 1.7151453495025635,
+      "learning_rate": 6.302968725956826e-07,
+      "loss": 0.1605,
+      "step": 64080
+    },
+    {
+      "epoch": 0.8435542454071695,
+      "grad_norm": 2.0342636108398438,
+      "learning_rate": 6.282268851471157e-07,
+      "loss": 0.1748,
+      "step": 64100
+    },
+    {
+      "epoch": 0.8438174448597147,
+      "grad_norm": 1.6327016353607178,
+      "learning_rate": 6.261600744853585e-07,
+      "loss": 0.1963,
+      "step": 64120
+    },
+    {
+      "epoch": 0.8440806443122598,
+      "grad_norm": 1.9226868152618408,
+      "learning_rate": 6.240964421122814e-07,
+      "loss": 0.1588,
+      "step": 64140
+    },
+    {
+      "epoch": 0.844343843764805,
+      "grad_norm": 1.8105432987213135,
+      "learning_rate": 6.220359895274358e-07,
+      "loss": 0.1766,
+      "step": 64160
+    },
+    {
+      "epoch": 0.8446070432173501,
+      "grad_norm": 1.510468602180481,
+      "learning_rate": 6.199787182280687e-07,
+      "loss": 0.1622,
+      "step": 64180
+    },
+    {
+      "epoch": 0.8448702426698953,
+      "grad_norm": 1.500399112701416,
+      "learning_rate": 6.179246297091157e-07,
+      "loss": 0.1571,
+      "step": 64200
+    },
+    {
+      "epoch": 0.8451334421224403,
+      "grad_norm": 1.5340521335601807,
+      "learning_rate": 6.158737254631969e-07,
+      "loss": 0.1672,
+      "step": 64220
+    },
+    {
+      "epoch": 0.8453966415749855,
+      "grad_norm": 1.586969256401062,
+      "learning_rate": 6.138260069806202e-07,
+      "loss": 0.1661,
+      "step": 64240
+    },
+    {
+      "epoch": 0.8456598410275307,
+      "grad_norm": 1.5701420307159424,
+      "learning_rate": 6.117814757493773e-07,
+      "loss": 0.1722,
+      "step": 64260
+    },
+    {
+      "epoch": 0.8459230404800758,
+      "grad_norm": 1.5603119134902954,
+      "learning_rate": 6.09740133255144e-07,
+      "loss": 0.1688,
+      "step": 64280
+    },
+    {
+      "epoch": 0.846186239932621,
+      "grad_norm": 1.8161264657974243,
+      "learning_rate": 6.077019809812807e-07,
+      "loss": 0.1705,
+      "step": 64300
+    },
+    {
+      "epoch": 0.846449439385166,
+      "grad_norm": 1.388864278793335,
+      "learning_rate": 6.056670204088288e-07,
+      "loss": 0.1667,
+      "step": 64320
+    },
+    {
+      "epoch": 0.8467126388377112,
+      "grad_norm": 1.5631577968597412,
+      "learning_rate": 6.03635253016509e-07,
+      "loss": 0.1573,
+      "step": 64340
+    },
+    {
+      "epoch": 0.8469758382902564,
+      "grad_norm": 1.3315999507904053,
+      "learning_rate": 6.01606680280723e-07,
+      "loss": 0.1722,
+      "step": 64360
+    },
+    {
+      "epoch": 0.8472390377428015,
+      "grad_norm": 1.644612193107605,
+      "learning_rate": 5.995813036755505e-07,
+      "loss": 0.1861,
+      "step": 64380
+    },
+    {
+      "epoch": 0.8475022371953467,
+      "grad_norm": 1.592258334159851,
+      "learning_rate": 5.975591246727496e-07,
+      "loss": 0.1733,
+      "step": 64400
+    },
+    {
+      "epoch": 0.8477654366478917,
+      "grad_norm": 1.2154418230056763,
+      "learning_rate": 5.955401447417536e-07,
+      "loss": 0.1782,
+      "step": 64420
+    },
+    {
+      "epoch": 0.8480286361004369,
+      "grad_norm": 1.402616262435913,
+      "learning_rate": 5.935243653496703e-07,
+      "loss": 0.1656,
+      "step": 64440
+    },
+    {
+      "epoch": 0.848291835552982,
+      "grad_norm": 1.9900575876235962,
+      "learning_rate": 5.915117879612864e-07,
+      "loss": 0.1721,
+      "step": 64460
+    },
+    {
+      "epoch": 0.8485550350055272,
+      "grad_norm": 1.4161818027496338,
+      "learning_rate": 5.89502414039057e-07,
+      "loss": 0.1456,
+      "step": 64480
+    },
+    {
+      "epoch": 0.8488182344580724,
+      "grad_norm": 1.8913651704788208,
+      "learning_rate": 5.874962450431116e-07,
+      "loss": 0.177,
+      "step": 64500
+    },
+    {
+      "epoch": 0.8488182344580724,
+      "eval_loss": 0.21353697776794434,
+      "eval_runtime": 4.6118,
+      "eval_samples_per_second": 3.469,
+      "eval_steps_per_second": 1.735,
+      "step": 64500
+    },
+    {
+      "epoch": 0.8490814339106174,
+      "grad_norm": 1.5366932153701782,
+      "learning_rate": 5.854932824312498e-07,
+      "loss": 0.2009,
+      "step": 64520
+    },
+    {
+      "epoch": 0.8493446333631626,
+      "grad_norm": 2.2525200843811035,
+      "learning_rate": 5.834935276589409e-07,
+      "loss": 0.1644,
+      "step": 64540
+    },
+    {
+      "epoch": 0.8496078328157077,
+      "grad_norm": 1.6828058958053589,
+      "learning_rate": 5.814969821793271e-07,
+      "loss": 0.1646,
+      "step": 64560
+    },
+    {
+      "epoch": 0.8498710322682529,
+      "grad_norm": 1.9800013303756714,
+      "learning_rate": 5.795036474432109e-07,
+      "loss": 0.1863,
+      "step": 64580
+    },
+    {
+      "epoch": 0.850134231720798,
+      "grad_norm": 1.6769779920578003,
+      "learning_rate": 5.775135248990699e-07,
+      "loss": 0.1729,
+      "step": 64600
+    },
+    {
+      "epoch": 0.8503974311733432,
+      "grad_norm": 2.0682051181793213,
+      "learning_rate": 5.755266159930428e-07,
+      "loss": 0.1811,
+      "step": 64620
+    },
+    {
+      "epoch": 0.8506606306258883,
+      "grad_norm": 2.0947020053863525,
+      "learning_rate": 5.735429221689337e-07,
+      "loss": 0.1866,
+      "step": 64640
+    },
+    {
+      "epoch": 0.8509238300784334,
+      "grad_norm": 1.620220422744751,
+      "learning_rate": 5.715624448682116e-07,
+      "loss": 0.1554,
+      "step": 64660
+    },
+    {
+      "epoch": 0.8511870295309786,
+      "grad_norm": 1.1789395809173584,
+      "learning_rate": 5.69585185530006e-07,
+      "loss": 0.1668,
+      "step": 64680
+    },
+    {
+      "epoch": 0.8514502289835237,
+      "grad_norm": 1.0477930307388306,
+      "learning_rate": 5.676111455911121e-07,
+      "loss": 0.1858,
+      "step": 64700
+    },
+    {
+      "epoch": 0.8517134284360689,
+      "grad_norm": 1.314913034439087,
+      "learning_rate": 5.656403264859828e-07,
+      "loss": 0.162,
+      "step": 64720
+    },
+    {
+      "epoch": 0.851976627888614,
+      "grad_norm": 1.7257815599441528,
+      "learning_rate": 5.636727296467276e-07,
+      "loss": 0.1582,
+      "step": 64740
+    },
+    {
+      "epoch": 0.8522398273411591,
+      "grad_norm": 1.773746371269226,
+      "learning_rate": 5.617083565031206e-07,
+      "loss": 0.1633,
+      "step": 64760
+    },
+    {
+      "epoch": 0.8525030267937043,
+      "grad_norm": 1.5452054738998413,
+      "learning_rate": 5.597472084825905e-07,
+      "loss": 0.1664,
+      "step": 64780
+    },
+    {
+      "epoch": 0.8527662262462494,
+      "grad_norm": 1.6670055389404297,
+      "learning_rate": 5.577892870102214e-07,
+      "loss": 0.1726,
+      "step": 64800
+    },
+    {
+      "epoch": 0.8530294256987946,
+      "grad_norm": 1.78778874874115,
+      "learning_rate": 5.55834593508755e-07,
+      "loss": 0.1726,
+      "step": 64820
+    },
+    {
+      "epoch": 0.8532926251513396,
+      "grad_norm": 1.5810613632202148,
+      "learning_rate": 5.538831293985847e-07,
+      "loss": 0.17,
+      "step": 64840
+    },
+    {
+      "epoch": 0.8535558246038848,
+      "grad_norm": 1.1657823324203491,
+      "learning_rate": 5.519348960977605e-07,
+      "loss": 0.1692,
+      "step": 64860
+    },
+    {
+      "epoch": 0.85381902405643,
+      "grad_norm": 1.6638901233673096,
+      "learning_rate": 5.499898950219823e-07,
+      "loss": 0.1525,
+      "step": 64880
+    },
+    {
+      "epoch": 0.8540822235089751,
+      "grad_norm": 1.7696946859359741,
+      "learning_rate": 5.480481275846022e-07,
+      "loss": 0.1808,
+      "step": 64900
+    },
+    {
+      "epoch": 0.8543454229615203,
+      "grad_norm": 2.5682170391082764,
+      "learning_rate": 5.461095951966222e-07,
+      "loss": 0.1747,
+      "step": 64920
+    },
+    {
+      "epoch": 0.8546086224140654,
+      "grad_norm": 1.7614803314208984,
+      "learning_rate": 5.441742992666926e-07,
+      "loss": 0.168,
+      "step": 64940
+    },
+    {
+      "epoch": 0.8548718218666105,
+      "grad_norm": 1.787857174873352,
+      "learning_rate": 5.422422412011158e-07,
+      "loss": 0.1545,
+      "step": 64960
+    },
+    {
+      "epoch": 0.8551350213191556,
+      "grad_norm": 1.4425028562545776,
+      "learning_rate": 5.403134224038359e-07,
+      "loss": 0.1647,
+      "step": 64980
+    },
+    {
+      "epoch": 0.8553982207717008,
+      "grad_norm": 1.826051115989685,
+      "learning_rate": 5.383878442764457e-07,
+      "loss": 0.1641,
+      "step": 65000
+    },
+    {
+      "epoch": 0.8553982207717008,
+      "eval_loss": 0.19906893372535706,
+      "eval_runtime": 4.4557,
+      "eval_samples_per_second": 3.591,
+      "eval_steps_per_second": 1.795,
+      "step": 65000
+    },
+    {
+      "epoch": 0.855661420224246,
+      "grad_norm": 2.133490800857544,
+      "learning_rate": 5.364655082181847e-07,
+      "loss": 0.1707,
+      "step": 65020
+    },
+    {
+      "epoch": 0.8559246196767911,
+      "grad_norm": 1.9316006898880005,
+      "learning_rate": 5.345464156259351e-07,
+      "loss": 0.1648,
+      "step": 65040
+    },
+    {
+      "epoch": 0.8561878191293362,
+      "grad_norm": 1.4846433401107788,
+      "learning_rate": 5.326305678942212e-07,
+      "loss": 0.181,
+      "step": 65060
+    },
+    {
+      "epoch": 0.8564510185818813,
+      "grad_norm": 1.653512954711914,
+      "learning_rate": 5.307179664152107e-07,
+      "loss": 0.1676,
+      "step": 65080
+    },
+    {
+      "epoch": 0.8567142180344265,
+      "grad_norm": 1.799439787864685,
+      "learning_rate": 5.288086125787106e-07,
+      "loss": 0.1561,
+      "step": 65100
+    },
+    {
+      "epoch": 0.8569774174869716,
+      "grad_norm": 1.417400598526001,
+      "learning_rate": 5.269025077721724e-07,
+      "loss": 0.1697,
+      "step": 65120
+    },
+    {
+      "epoch": 0.8572406169395168,
+      "grad_norm": 1.6313802003860474,
+      "learning_rate": 5.249996533806828e-07,
+      "loss": 0.1525,
+      "step": 65140
+    },
+    {
+      "epoch": 0.8575038163920619,
+      "grad_norm": 1.7196242809295654,
+      "learning_rate": 5.23100050786965e-07,
+      "loss": 0.165,
+      "step": 65160
+    },
+    {
+      "epoch": 0.857767015844607,
+      "grad_norm": 2.0277724266052246,
+      "learning_rate": 5.212037013713839e-07,
+      "loss": 0.1735,
+      "step": 65180
+    },
+    {
+      "epoch": 0.8580302152971522,
+      "grad_norm": 1.691499948501587,
+      "learning_rate": 5.193106065119369e-07,
+      "loss": 0.1701,
+      "step": 65200
+    },
+    {
+      "epoch": 0.8582934147496973,
+      "grad_norm": 1.5974180698394775,
+      "learning_rate": 5.174207675842607e-07,
+      "loss": 0.1667,
+      "step": 65220
+    },
+    {
+      "epoch": 0.8585566142022425,
+      "grad_norm": 1.8551092147827148,
+      "learning_rate": 5.155341859616203e-07,
+      "loss": 0.1626,
+      "step": 65240
+    },
+    {
+      "epoch": 0.8588198136547875,
+      "grad_norm": 1.5553995370864868,
+      "learning_rate": 5.136508630149161e-07,
+      "loss": 0.1809,
+      "step": 65260
+    },
+    {
+      "epoch": 0.8590830131073327,
+      "grad_norm": 2.2132177352905273,
+      "learning_rate": 5.117708001126836e-07,
+      "loss": 0.1771,
+      "step": 65280
+    },
+    {
+      "epoch": 0.8593462125598779,
+      "grad_norm": 2.070216655731201,
+      "learning_rate": 5.098939986210854e-07,
+      "loss": 0.1838,
+      "step": 65300
+    },
+    {
+      "epoch": 0.859609412012423,
+      "grad_norm": 1.4257506132125854,
+      "learning_rate": 5.080204599039152e-07,
+      "loss": 0.1466,
+      "step": 65320
+    },
+    {
+      "epoch": 0.8598726114649682,
+      "grad_norm": 1.8143725395202637,
+      "learning_rate": 5.06150185322597e-07,
+      "loss": 0.1804,
+      "step": 65340
+    },
+    {
+      "epoch": 0.8601358109175133,
+      "grad_norm": 1.312599539756775,
+      "learning_rate": 5.042831762361805e-07,
+      "loss": 0.1653,
+      "step": 65360
+    },
+    {
+      "epoch": 0.8603990103700584,
+      "grad_norm": 1.6162930727005005,
+      "learning_rate": 5.024194340013471e-07,
+      "loss": 0.1652,
+      "step": 65380
+    },
+    {
+      "epoch": 0.8606622098226036,
+      "grad_norm": 1.1971834897994995,
+      "learning_rate": 5.005589599723981e-07,
+      "loss": 0.1624,
+      "step": 65400
+    },
+    {
+      "epoch": 0.8609254092751487,
+      "grad_norm": 1.7098243236541748,
+      "learning_rate": 4.987017555012635e-07,
+      "loss": 0.1711,
+      "step": 65420
+    },
+    {
+      "epoch": 0.8611886087276939,
+      "grad_norm": 1.7987123727798462,
+      "learning_rate": 4.968478219374984e-07,
+      "loss": 0.1769,
+      "step": 65440
+    },
+    {
+      "epoch": 0.861451808180239,
+      "grad_norm": 1.5199477672576904,
+      "learning_rate": 4.949971606282794e-07,
+      "loss": 0.1644,
+      "step": 65460
+    },
+    {
+      "epoch": 0.8617150076327841,
+      "grad_norm": 1.5326142311096191,
+      "learning_rate": 4.931497729184053e-07,
+      "loss": 0.1708,
+      "step": 65480
+    },
+    {
+      "epoch": 0.8619782070853292,
+      "grad_norm": 1.429611086845398,
+      "learning_rate": 4.913056601502963e-07,
+      "loss": 0.1782,
+      "step": 65500
+    },
+    {
+      "epoch": 0.8619782070853292,
+      "eval_loss": 0.20205625891685486,
+      "eval_runtime": 4.4104,
+      "eval_samples_per_second": 3.628,
+      "eval_steps_per_second": 1.814,
+      "step": 65500
+    },
+    {
+      "epoch": 0.8622414065378744,
+      "grad_norm": 1.7161164283752441,
+      "learning_rate": 4.894648236639915e-07,
+      "loss": 0.1728,
+      "step": 65520
+    },
+    {
+      "epoch": 0.8625046059904196,
+      "grad_norm": 2.105393886566162,
+      "learning_rate": 4.876272647971536e-07,
+      "loss": 0.1724,
+      "step": 65540
+    },
+    {
+      "epoch": 0.8627678054429647,
+      "grad_norm": 1.7519155740737915,
+      "learning_rate": 4.857929848850601e-07,
+      "loss": 0.1692,
+      "step": 65560
+    },
+    {
+      "epoch": 0.8630310048955098,
+      "grad_norm": 2.2599992752075195,
+      "learning_rate": 4.839619852606037e-07,
+      "loss": 0.1736,
+      "step": 65580
+    },
+    {
+      "epoch": 0.8632942043480549,
+      "grad_norm": 1.523550271987915,
+      "learning_rate": 4.821342672542984e-07,
+      "loss": 0.1616,
+      "step": 65600
+    },
+    {
+      "epoch": 0.8635574038006001,
+      "grad_norm": 1.347743034362793,
+      "learning_rate": 4.803098321942712e-07,
+      "loss": 0.1551,
+      "step": 65620
+    },
+    {
+      "epoch": 0.8638206032531452,
+      "grad_norm": 2.027175188064575,
+      "learning_rate": 4.784886814062628e-07,
+      "loss": 0.1798,
+      "step": 65640
+    },
+    {
+      "epoch": 0.8640838027056904,
+      "grad_norm": 1.6077977418899536,
+      "learning_rate": 4.766708162136291e-07,
+      "loss": 0.169,
+      "step": 65660
+    },
+    {
+      "epoch": 0.8643470021582355,
+      "grad_norm": 1.6331498622894287,
+      "learning_rate": 4.7485623793733615e-07,
+      "loss": 0.1523,
+      "step": 65680
+    },
+    {
+      "epoch": 0.8646102016107806,
+      "grad_norm": 1.8278828859329224,
+      "learning_rate": 4.7304494789596536e-07,
+      "loss": 0.1688,
+      "step": 65700
+    },
+    {
+      "epoch": 0.8648734010633258,
+      "grad_norm": 1.2686405181884766,
+      "learning_rate": 4.7123694740570434e-07,
+      "loss": 0.1644,
+      "step": 65720
+    },
+    {
+      "epoch": 0.8651366005158709,
+      "grad_norm": 2.1564438343048096,
+      "learning_rate": 4.694322377803534e-07,
+      "loss": 0.1709,
+      "step": 65740
+    },
+    {
+      "epoch": 0.8653997999684161,
+      "grad_norm": 1.809579610824585,
+      "learning_rate": 4.6763082033131955e-07,
+      "loss": 0.1693,
+      "step": 65760
+    },
+    {
+      "epoch": 0.8656629994209613,
+      "grad_norm": 1.2358664274215698,
+      "learning_rate": 4.65832696367618e-07,
+      "loss": 0.1451,
+      "step": 65780
+    },
+    {
+      "epoch": 0.8659261988735063,
+      "grad_norm": 1.3446969985961914,
+      "learning_rate": 4.6403786719587394e-07,
+      "loss": 0.1766,
+      "step": 65800
+    },
+    {
+      "epoch": 0.8661893983260515,
+      "grad_norm": 1.8859080076217651,
+      "learning_rate": 4.622463341203126e-07,
+      "loss": 0.1632,
+      "step": 65820
+    },
+    {
+      "epoch": 0.8664525977785966,
+      "grad_norm": 2.1976218223571777,
+      "learning_rate": 4.60458098442767e-07,
+      "loss": 0.1792,
+      "step": 65840
+    },
+    {
+      "epoch": 0.8667157972311418,
+      "grad_norm": 1.6627578735351562,
+      "learning_rate": 4.58673161462676e-07,
+      "loss": 0.1702,
+      "step": 65860
+    },
+    {
+      "epoch": 0.8669789966836869,
+      "grad_norm": 1.9740493297576904,
+      "learning_rate": 4.568915244770783e-07,
+      "loss": 0.1451,
+      "step": 65880
+    },
+    {
+      "epoch": 0.867242196136232,
+      "grad_norm": 1.4959428310394287,
+      "learning_rate": 4.551131887806165e-07,
+      "loss": 0.1765,
+      "step": 65900
+    },
+    {
+      "epoch": 0.8675053955887772,
+      "grad_norm": 2.5356688499450684,
+      "learning_rate": 4.533381556655325e-07,
+      "loss": 0.165,
+      "step": 65920
+    },
+    {
+      "epoch": 0.8677685950413223,
+      "grad_norm": 1.6718651056289673,
+      "learning_rate": 4.515664264216701e-07,
+      "loss": 0.1684,
+      "step": 65940
+    },
+    {
+      "epoch": 0.8680317944938675,
+      "grad_norm": 2.1584017276763916,
+      "learning_rate": 4.497980023364723e-07,
+      "loss": 0.1983,
+      "step": 65960
+    },
+    {
+      "epoch": 0.8682949939464126,
+      "grad_norm": 2.034057140350342,
+      "learning_rate": 4.480328846949794e-07,
+      "loss": 0.1715,
+      "step": 65980
+    },
+    {
+      "epoch": 0.8685581933989577,
+      "grad_norm": 1.308982014656067,
+      "learning_rate": 4.462710747798293e-07,
+      "loss": 0.183,
+      "step": 66000
+    },
+    {
+      "epoch": 0.8685581933989577,
+      "eval_loss": 0.1967773139476776,
+      "eval_runtime": 4.7184,
+      "eval_samples_per_second": 3.391,
+      "eval_steps_per_second": 1.695,
+      "step": 66000
+    },
+    {
+      "epoch": 0.8688213928515028,
+      "grad_norm": 1.2660967111587524,
+      "learning_rate": 4.4451257387125636e-07,
+      "loss": 0.1623,
+      "step": 66020
+    },
+    {
+      "epoch": 0.869084592304048,
+      "grad_norm": 1.836686372756958,
+      "learning_rate": 4.4275738324709106e-07,
+      "loss": 0.1558,
+      "step": 66040
+    },
+    {
+      "epoch": 0.8693477917565932,
+      "grad_norm": 2.338085889816284,
+      "learning_rate": 4.410055041827571e-07,
+      "loss": 0.1746,
+      "step": 66060
+    },
+    {
+      "epoch": 0.8696109912091383,
+      "grad_norm": 1.2496567964553833,
+      "learning_rate": 4.3925693795127177e-07,
+      "loss": 0.1505,
+      "step": 66080
+    },
+    {
+      "epoch": 0.8698741906616835,
+      "grad_norm": 1.8931291103363037,
+      "learning_rate": 4.3751168582324866e-07,
+      "loss": 0.1579,
+      "step": 66100
+    },
+    {
+      "epoch": 0.8701373901142285,
+      "grad_norm": 1.9000297784805298,
+      "learning_rate": 4.3576974906688774e-07,
+      "loss": 0.1849,
+      "step": 66120
+    },
+    {
+      "epoch": 0.8704005895667737,
+      "grad_norm": 1.2805622816085815,
+      "learning_rate": 4.340311289479843e-07,
+      "loss": 0.1845,
+      "step": 66140
+    },
+    {
+      "epoch": 0.8706637890193188,
+      "grad_norm": 1.671220064163208,
+      "learning_rate": 4.3229582672992045e-07,
+      "loss": 0.1757,
+      "step": 66160
+    },
+    {
+      "epoch": 0.870926988471864,
+      "grad_norm": 1.9813323020935059,
+      "learning_rate": 4.305638436736687e-07,
+      "loss": 0.1779,
+      "step": 66180
+    },
+    {
+      "epoch": 0.8711901879244092,
+      "grad_norm": 2.4355337619781494,
+      "learning_rate": 4.2883518103778943e-07,
+      "loss": 0.1714,
+      "step": 66200
+    },
+    {
+      "epoch": 0.8714533873769542,
+      "grad_norm": 1.7350873947143555,
+      "learning_rate": 4.2710984007843183e-07,
+      "loss": 0.1689,
+      "step": 66220
+    },
+    {
+      "epoch": 0.8717165868294994,
+      "grad_norm": 1.2526566982269287,
+      "learning_rate": 4.253878220493274e-07,
+      "loss": 0.1643,
+      "step": 66240
+    },
+    {
+      "epoch": 0.8719797862820445,
+      "grad_norm": 3.3615050315856934,
+      "learning_rate": 4.2366912820179797e-07,
+      "loss": 0.1743,
+      "step": 66260
+    },
+    {
+      "epoch": 0.8722429857345897,
+      "grad_norm": 1.7450515031814575,
+      "learning_rate": 4.219537597847456e-07,
+      "loss": 0.178,
+      "step": 66280
+    },
+    {
+      "epoch": 0.8725061851871349,
+      "grad_norm": 1.6590707302093506,
+      "learning_rate": 4.2024171804465827e-07,
+      "loss": 0.17,
+      "step": 66300
+    },
+    {
+      "epoch": 0.8727693846396799,
+      "grad_norm": 2.441903829574585,
+      "learning_rate": 4.185330042256058e-07,
+      "loss": 0.1548,
+      "step": 66320
+    },
+    {
+      "epoch": 0.8730325840922251,
+      "grad_norm": 1.5902758836746216,
+      "learning_rate": 4.168276195692389e-07,
+      "loss": 0.1634,
+      "step": 66340
+    },
+    {
+      "epoch": 0.8732957835447702,
+      "grad_norm": 1.4393171072006226,
+      "learning_rate": 4.151255653147923e-07,
+      "loss": 0.1833,
+      "step": 66360
+    },
+    {
+      "epoch": 0.8735589829973154,
+      "grad_norm": 1.4261935949325562,
+      "learning_rate": 4.1342684269907783e-07,
+      "loss": 0.1822,
+      "step": 66380
+    },
+    {
+      "epoch": 0.8738221824498605,
+      "grad_norm": 1.6420830488204956,
+      "learning_rate": 4.1173145295648477e-07,
+      "loss": 0.1713,
+      "step": 66400
+    },
+    {
+      "epoch": 0.8740853819024056,
+      "grad_norm": 1.556566834449768,
+      "learning_rate": 4.100393973189853e-07,
+      "loss": 0.1844,
+      "step": 66420
+    },
+    {
+      "epoch": 0.8743485813549507,
+      "grad_norm": 1.7923128604888916,
+      "learning_rate": 4.0835067701612554e-07,
+      "loss": 0.1659,
+      "step": 66440
+    },
+    {
+      "epoch": 0.8746117808074959,
+      "grad_norm": 1.9670671224594116,
+      "learning_rate": 4.066652932750287e-07,
+      "loss": 0.1611,
+      "step": 66460
+    },
+    {
+      "epoch": 0.8748749802600411,
+      "grad_norm": 2.4062581062316895,
+      "learning_rate": 4.0498324732039354e-07,
+      "loss": 0.1702,
+      "step": 66480
+    },
+    {
+      "epoch": 0.8751381797125862,
+      "grad_norm": 1.2815577983856201,
+      "learning_rate": 4.033045403744923e-07,
+      "loss": 0.1735,
+      "step": 66500
+    },
+    {
+      "epoch": 0.8751381797125862,
+      "eval_loss": 0.2077566683292389,
+      "eval_runtime": 4.4495,
+      "eval_samples_per_second": 3.596,
+      "eval_steps_per_second": 1.798,
+      "step": 66500
+    },
+    {
+      "epoch": 0.8754013791651314,
+      "grad_norm": 1.3562498092651367,
+      "learning_rate": 4.0162917365717315e-07,
+      "loss": 0.17,
+      "step": 66520
+    },
+    {
+      "epoch": 0.8756645786176764,
+      "grad_norm": 1.2158669233322144,
+      "learning_rate": 3.999571483858561e-07,
+      "loss": 0.1687,
+      "step": 66540
+    },
+    {
+      "epoch": 0.8759277780702216,
+      "grad_norm": 1.724364995956421,
+      "learning_rate": 3.9828846577553203e-07,
+      "loss": 0.1838,
+      "step": 66560
+    },
+    {
+      "epoch": 0.8761909775227668,
+      "grad_norm": 1.4473729133605957,
+      "learning_rate": 3.966231270387638e-07,
+      "loss": 0.1891,
+      "step": 66580
+    },
+    {
+      "epoch": 0.8764541769753119,
+      "grad_norm": 1.4290999174118042,
+      "learning_rate": 3.9496113338568345e-07,
+      "loss": 0.1791,
+      "step": 66600
+    },
+    {
+      "epoch": 0.8767173764278571,
+      "grad_norm": 1.653855800628662,
+      "learning_rate": 3.9330248602399536e-07,
+      "loss": 0.181,
+      "step": 66620
+    },
+    {
+      "epoch": 0.8769805758804021,
+      "grad_norm": 1.8037916421890259,
+      "learning_rate": 3.916471861589677e-07,
+      "loss": 0.1731,
+      "step": 66640
+    },
+    {
+      "epoch": 0.8772437753329473,
+      "grad_norm": 1.5368127822875977,
+      "learning_rate": 3.899952349934383e-07,
+      "loss": 0.1657,
+      "step": 66660
+    },
+    {
+      "epoch": 0.8775069747854924,
+      "grad_norm": 1.5534662008285522,
+      "learning_rate": 3.8834663372781356e-07,
+      "loss": 0.1686,
+      "step": 66680
+    },
+    {
+      "epoch": 0.8777701742380376,
+      "grad_norm": 1.5343372821807861,
+      "learning_rate": 3.8670138356006303e-07,
+      "loss": 0.1718,
+      "step": 66700
+    },
+    {
+      "epoch": 0.8780333736905828,
+      "grad_norm": 2.3240718841552734,
+      "learning_rate": 3.850594856857226e-07,
+      "loss": 0.1941,
+      "step": 66720
+    },
+    {
+      "epoch": 0.8782965731431278,
+      "grad_norm": 1.2541682720184326,
+      "learning_rate": 3.834209412978901e-07,
+      "loss": 0.1659,
+      "step": 66740
+    },
+    {
+      "epoch": 0.878559772595673,
+      "grad_norm": 1.2523822784423828,
+      "learning_rate": 3.8178575158722864e-07,
+      "loss": 0.1766,
+      "step": 66760
+    },
+    {
+      "epoch": 0.8788229720482181,
+      "grad_norm": 1.4435845613479614,
+      "learning_rate": 3.8015391774196443e-07,
+      "loss": 0.1785,
+      "step": 66780
+    },
+    {
+      "epoch": 0.8790861715007633,
+      "grad_norm": 1.8097612857818604,
+      "learning_rate": 3.7852544094788277e-07,
+      "loss": 0.1734,
+      "step": 66800
+    },
+    {
+      "epoch": 0.8793493709533085,
+      "grad_norm": 1.7031244039535522,
+      "learning_rate": 3.769003223883283e-07,
+      "loss": 0.189,
+      "step": 66820
+    },
+    {
+      "epoch": 0.8796125704058536,
+      "grad_norm": 1.4425146579742432,
+      "learning_rate": 3.7527856324421075e-07,
+      "loss": 0.1698,
+      "step": 66840
+    },
+    {
+      "epoch": 0.8798757698583987,
+      "grad_norm": 1.6217784881591797,
+      "learning_rate": 3.7366016469399303e-07,
+      "loss": 0.1874,
+      "step": 66860
+    },
+    {
+      "epoch": 0.8801389693109438,
+      "grad_norm": 1.8848726749420166,
+      "learning_rate": 3.7204512791370106e-07,
+      "loss": 0.1775,
+      "step": 66880
+    },
+    {
+      "epoch": 0.880402168763489,
+      "grad_norm": 1.4746283292770386,
+      "learning_rate": 3.7043345407691334e-07,
+      "loss": 0.1695,
+      "step": 66900
+    },
+    {
+      "epoch": 0.8806653682160341,
+      "grad_norm": 1.6697168350219727,
+      "learning_rate": 3.688251443547658e-07,
+      "loss": 0.1685,
+      "step": 66920
+    },
+    {
+      "epoch": 0.8809285676685793,
+      "grad_norm": 1.324507236480713,
+      "learning_rate": 3.672201999159525e-07,
+      "loss": 0.1833,
+      "step": 66940
+    },
+    {
+      "epoch": 0.8811917671211243,
+      "grad_norm": 1.5267077684402466,
+      "learning_rate": 3.6561862192671994e-07,
+      "loss": 0.1578,
+      "step": 66960
+    },
+    {
+      "epoch": 0.8814549665736695,
+      "grad_norm": 1.5439190864562988,
+      "learning_rate": 3.640204115508683e-07,
+      "loss": 0.1862,
+      "step": 66980
+    },
+    {
+      "epoch": 0.8817181660262147,
+      "grad_norm": 1.784958004951477,
+      "learning_rate": 3.6242556994975143e-07,
+      "loss": 0.1588,
+      "step": 67000
+    },
+    {
+      "epoch": 0.8817181660262147,
+      "eval_loss": 0.19462963938713074,
+      "eval_runtime": 2.9641,
+      "eval_samples_per_second": 5.398,
+      "eval_steps_per_second": 2.699,
+      "step": 67000
+    },
+    {
+      "epoch": 0.8819813654787598,
+      "grad_norm": 1.7258950471878052,
+      "learning_rate": 3.608340982822728e-07,
+      "loss": 0.1838,
+      "step": 67020
+    },
+    {
+      "epoch": 0.882244564931305,
+      "grad_norm": 1.3943952322006226,
+      "learning_rate": 3.592459977048923e-07,
+      "loss": 0.1702,
+      "step": 67040
+    },
+    {
+      "epoch": 0.88250776438385,
+      "grad_norm": 1.4951978921890259,
+      "learning_rate": 3.5766126937161526e-07,
+      "loss": 0.171,
+      "step": 67060
+    },
+    {
+      "epoch": 0.8827709638363952,
+      "grad_norm": 1.5643328428268433,
+      "learning_rate": 3.560799144339966e-07,
+      "loss": 0.1575,
+      "step": 67080
+    },
+    {
+      "epoch": 0.8830341632889404,
+      "grad_norm": 1.6817100048065186,
+      "learning_rate": 3.5450193404114487e-07,
+      "loss": 0.1577,
+      "step": 67100
+    },
+    {
+      "epoch": 0.8832973627414855,
+      "grad_norm": 1.9669744968414307,
+      "learning_rate": 3.5292732933971116e-07,
+      "loss": 0.1641,
+      "step": 67120
+    },
+    {
+      "epoch": 0.8835605621940307,
+      "grad_norm": 1.891140103340149,
+      "learning_rate": 3.5135610147389644e-07,
+      "loss": 0.1772,
+      "step": 67140
+    },
+    {
+      "epoch": 0.8838237616465757,
+      "grad_norm": 1.622894048690796,
+      "learning_rate": 3.4978825158544616e-07,
+      "loss": 0.1589,
+      "step": 67160
+    },
+    {
+      "epoch": 0.8840869610991209,
+      "grad_norm": 1.5769035816192627,
+      "learning_rate": 3.4822378081365236e-07,
+      "loss": 0.1676,
+      "step": 67180
+    },
+    {
+      "epoch": 0.884350160551666,
+      "grad_norm": 1.4430614709854126,
+      "learning_rate": 3.4666269029535184e-07,
+      "loss": 0.1694,
+      "step": 67200
+    },
+    {
+      "epoch": 0.8846133600042112,
+      "grad_norm": 1.553456425666809,
+      "learning_rate": 3.451049811649254e-07,
+      "loss": 0.1626,
+      "step": 67220
+    },
+    {
+      "epoch": 0.8848765594567564,
+      "grad_norm": 1.7416012287139893,
+      "learning_rate": 3.435506545542927e-07,
+      "loss": 0.1789,
+      "step": 67240
+    },
+    {
+      "epoch": 0.8851397589093015,
+      "grad_norm": 1.6421476602554321,
+      "learning_rate": 3.4199971159292134e-07,
+      "loss": 0.1689,
+      "step": 67260
+    },
+    {
+      "epoch": 0.8854029583618466,
+      "grad_norm": 1.3852603435516357,
+      "learning_rate": 3.40452153407817e-07,
+      "loss": 0.158,
+      "step": 67280
+    },
+    {
+      "epoch": 0.8856661578143917,
+      "grad_norm": 1.6317458152770996,
+      "learning_rate": 3.389079811235263e-07,
+      "loss": 0.1676,
+      "step": 67300
+    },
+    {
+      "epoch": 0.8859293572669369,
+      "grad_norm": 2.250880479812622,
+      "learning_rate": 3.373671958621344e-07,
+      "loss": 0.1749,
+      "step": 67320
+    },
+    {
+      "epoch": 0.886192556719482,
+      "grad_norm": 1.7971525192260742,
+      "learning_rate": 3.358297987432668e-07,
+      "loss": 0.1705,
+      "step": 67340
+    },
+    {
+      "epoch": 0.8864557561720272,
+      "grad_norm": 2.0014491081237793,
+      "learning_rate": 3.342957908840866e-07,
+      "loss": 0.1687,
+      "step": 67360
+    },
+    {
+      "epoch": 0.8867189556245723,
+      "grad_norm": 2.4138903617858887,
+      "learning_rate": 3.3276517339929424e-07,
+      "loss": 0.1686,
+      "step": 67380
+    },
+    {
+      "epoch": 0.8869821550771174,
+      "grad_norm": 1.5627574920654297,
+      "learning_rate": 3.3123794740112623e-07,
+      "loss": 0.1665,
+      "step": 67400
+    },
+    {
+      "epoch": 0.8872453545296626,
+      "grad_norm": 1.8441654443740845,
+      "learning_rate": 3.297141139993537e-07,
+      "loss": 0.1834,
+      "step": 67420
+    },
+    {
+      "epoch": 0.8875085539822077,
+      "grad_norm": 1.6202750205993652,
+      "learning_rate": 3.281936743012831e-07,
+      "loss": 0.1827,
+      "step": 67440
+    },
+    {
+      "epoch": 0.8877717534347529,
+      "grad_norm": 1.6369091272354126,
+      "learning_rate": 3.266766294117579e-07,
+      "loss": 0.1669,
+      "step": 67460
+    },
+    {
+      "epoch": 0.888034952887298,
+      "grad_norm": 1.700462818145752,
+      "learning_rate": 3.2516298043314797e-07,
+      "loss": 0.1657,
+      "step": 67480
+    },
+    {
+      "epoch": 0.8882981523398431,
+      "grad_norm": 2.1126701831817627,
+      "learning_rate": 3.2365272846536177e-07,
+      "loss": 0.1654,
+      "step": 67500
+    },
+    {
+      "epoch": 0.8882981523398431,
+      "eval_loss": 0.20013199746608734,
+      "eval_runtime": 4.0757,
+      "eval_samples_per_second": 3.926,
+      "eval_steps_per_second": 1.963,
+      "step": 67500
+    },
+    {
+      "epoch": 0.8885613517923883,
+      "grad_norm": 2.1401360034942627,
+      "learning_rate": 3.22145874605837e-07,
+      "loss": 0.1578,
+      "step": 67520
+    },
+    {
+      "epoch": 0.8888245512449334,
+      "grad_norm": 1.6239949464797974,
+      "learning_rate": 3.206424199495417e-07,
+      "loss": 0.1732,
+      "step": 67540
+    },
+    {
+      "epoch": 0.8890877506974786,
+      "grad_norm": 1.7708386182785034,
+      "learning_rate": 3.191423655889742e-07,
+      "loss": 0.1765,
+      "step": 67560
+    },
+    {
+      "epoch": 0.8893509501500236,
+      "grad_norm": 2.1668741703033447,
+      "learning_rate": 3.1764571261416135e-07,
+      "loss": 0.1637,
+      "step": 67580
+    },
+    {
+      "epoch": 0.8896141496025688,
+      "grad_norm": 1.659727931022644,
+      "learning_rate": 3.1615246211265937e-07,
+      "loss": 0.1917,
+      "step": 67600
+    },
+    {
+      "epoch": 0.889877349055114,
+      "grad_norm": 1.74148428440094,
+      "learning_rate": 3.1466261516955254e-07,
+      "loss": 0.1658,
+      "step": 67620
+    },
+    {
+      "epoch": 0.8901405485076591,
+      "grad_norm": 1.5698541402816772,
+      "learning_rate": 3.131761728674504e-07,
+      "loss": 0.163,
+      "step": 67640
+    },
+    {
+      "epoch": 0.8904037479602043,
+      "grad_norm": 1.72197425365448,
+      "learning_rate": 3.1169313628648947e-07,
+      "loss": 0.177,
+      "step": 67660
+    },
+    {
+      "epoch": 0.8906669474127494,
+      "grad_norm": 1.6280436515808105,
+      "learning_rate": 3.1021350650433044e-07,
+      "loss": 0.1583,
+      "step": 67680
+    },
+    {
+      "epoch": 0.8909301468652945,
+      "grad_norm": 1.7222532033920288,
+      "learning_rate": 3.0873728459615946e-07,
+      "loss": 0.1841,
+      "step": 67700
+    },
+    {
+      "epoch": 0.8911933463178396,
+      "grad_norm": 1.7605030536651611,
+      "learning_rate": 3.072644716346862e-07,
+      "loss": 0.1535,
+      "step": 67720
+    },
+    {
+      "epoch": 0.8914565457703848,
+      "grad_norm": 1.5122253894805908,
+      "learning_rate": 3.0579506869014233e-07,
+      "loss": 0.1804,
+      "step": 67740
+    },
+    {
+      "epoch": 0.89171974522293,
+      "grad_norm": 1.3153375387191772,
+      "learning_rate": 3.0432907683028256e-07,
+      "loss": 0.152,
+      "step": 67760
+    },
+    {
+      "epoch": 0.8919829446754751,
+      "grad_norm": 1.2126888036727905,
+      "learning_rate": 3.028664971203832e-07,
+      "loss": 0.1656,
+      "step": 67780
+    },
+    {
+      "epoch": 0.8922461441280202,
+      "grad_norm": 1.7752838134765625,
+      "learning_rate": 3.014073306232401e-07,
+      "loss": 0.1696,
+      "step": 67800
+    },
+    {
+      "epoch": 0.8925093435805653,
+      "grad_norm": 1.8188780546188354,
+      "learning_rate": 2.999515783991691e-07,
+      "loss": 0.1713,
+      "step": 67820
+    },
+    {
+      "epoch": 0.8927725430331105,
+      "grad_norm": 1.9752446413040161,
+      "learning_rate": 2.984992415060045e-07,
+      "loss": 0.1707,
+      "step": 67840
+    },
+    {
+      "epoch": 0.8930357424856556,
+      "grad_norm": 1.6852295398712158,
+      "learning_rate": 2.970503209990999e-07,
+      "loss": 0.1642,
+      "step": 67860
+    },
+    {
+      "epoch": 0.8932989419382008,
+      "grad_norm": 1.4846044778823853,
+      "learning_rate": 2.956048179313276e-07,
+      "loss": 0.1674,
+      "step": 67880
+    },
+    {
+      "epoch": 0.8935621413907459,
+      "grad_norm": 1.566523551940918,
+      "learning_rate": 2.9416273335307223e-07,
+      "loss": 0.1729,
+      "step": 67900
+    },
+    {
+      "epoch": 0.893825340843291,
+      "grad_norm": 1.4032057523727417,
+      "learning_rate": 2.9272406831223845e-07,
+      "loss": 0.1452,
+      "step": 67920
+    },
+    {
+      "epoch": 0.8940885402958362,
+      "grad_norm": 2.321786642074585,
+      "learning_rate": 2.9128882385424493e-07,
+      "loss": 0.1708,
+      "step": 67940
+    },
+    {
+      "epoch": 0.8943517397483813,
+      "grad_norm": 1.5116506814956665,
+      "learning_rate": 2.898570010220242e-07,
+      "loss": 0.1594,
+      "step": 67960
+    },
+    {
+      "epoch": 0.8946149392009265,
+      "grad_norm": 1.7108181715011597,
+      "learning_rate": 2.884286008560222e-07,
+      "loss": 0.1653,
+      "step": 67980
+    },
+    {
+      "epoch": 0.8948781386534717,
+      "grad_norm": 2.1078197956085205,
+      "learning_rate": 2.8700362439419815e-07,
+      "loss": 0.177,
+      "step": 68000
+    },
+    {
+      "epoch": 0.8948781386534717,
+      "eval_loss": 0.1999533772468567,
+      "eval_runtime": 4.4752,
+      "eval_samples_per_second": 3.575,
+      "eval_steps_per_second": 1.788,
+      "step": 68000
+    }
+  ],
+  "logging_steps": 20,
+  "max_steps": 75988,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.1493257886590566e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}